From ddf56714ad424c91c43da41aadd7201e52cb7952 Mon Sep 17 00:00:00 2001 From: leefer Date: Thu, 29 Oct 2026 22:00:41 +0800 Subject: [PATCH] HEL-892: bind supplied API companions with provenance and fail-closed import Co-authored-by: multica-agent --- .../scripts/audit_library_documents.py | 24 +- .../scripts/audit_supplied_api.py | 107 + .../src/calliope/comfyui/companions.py | 222 + .../src/calliope/comfyui/library.py | 51 +- .../src/calliope/comfyui/library_cli.py | 11 +- .../tests/test_supplied_companions.py | 229 + docs/hel-892/README.md | 2 + docs/hel-892/sixth-evidence/01.json | 2876 + docs/hel-892/sixth-evidence/02.json | 903 + docs/hel-892/sixth-evidence/03.json | 2211 + docs/hel-892/sixth-evidence/04.json | 3223 + docs/hel-892/sixth-evidence/05.json | 2887 + docs/hel-892/sixth-evidence/06.json | 6736 ++ docs/hel-892/sixth-evidence/07.json | 14095 +++++ .../hel-892/sixth-evidence/affected-tests.txt | 9 + .../sixth-evidence/backend-build-final.txt | 4 + docs/hel-892/sixth-evidence/changed-lint.txt | 1 + .../sixth-evidence/final-data-tests.txt | 8 + docs/hel-892/sixth-evidence/import-final.txt | 1 + docs/hel-892/sixth-evidence/paired-audit.txt | 1 + docs/hel-892/sixth-evidence/reimport.txt | 1 + docs/hel-892/sixth-evidence/source-audit.txt | 1 + .../sixth-evidence/source-document-audit.json | 51978 ++++++++++++++++ docs/hel-892/sixth-evidence/summary.json | 1003 + docs/hel-892/supplied-api/01.api.json | 304 + docs/hel-892/supplied-api/02.api.json | 86 + docs/hel-892/supplied-api/03.api.json | 186 + docs/hel-892/supplied-api/04.api.json | 244 + docs/hel-892/supplied-api/05.api.json | 250 + docs/hel-892/supplied-api/06.api.json | 457 + docs/hel-892/supplied-api/07.api.json | 1192 + docs/hel-892/supplied-api/manifest.json | 145 + docs/hel-892/第六批真实API配套导入.md | 64 + 33 files changed, 89504 insertions(+), 8 deletions(-) create mode 100644 calliope-backend/scripts/audit_supplied_api.py create mode 100644 calliope-backend/src/calliope/comfyui/companions.py create mode 100644 calliope-backend/tests/test_supplied_companions.py create mode 100644 docs/hel-892/sixth-evidence/01.json create mode 100644 docs/hel-892/sixth-evidence/02.json create mode 100644 docs/hel-892/sixth-evidence/03.json create mode 100644 docs/hel-892/sixth-evidence/04.json create mode 100644 docs/hel-892/sixth-evidence/05.json create mode 100644 docs/hel-892/sixth-evidence/06.json create mode 100644 docs/hel-892/sixth-evidence/07.json create mode 100644 docs/hel-892/sixth-evidence/affected-tests.txt create mode 100644 docs/hel-892/sixth-evidence/backend-build-final.txt create mode 100644 docs/hel-892/sixth-evidence/changed-lint.txt create mode 100644 docs/hel-892/sixth-evidence/final-data-tests.txt create mode 100644 docs/hel-892/sixth-evidence/import-final.txt create mode 100644 docs/hel-892/sixth-evidence/paired-audit.txt create mode 100644 docs/hel-892/sixth-evidence/reimport.txt create mode 100644 docs/hel-892/sixth-evidence/source-audit.txt create mode 100644 docs/hel-892/sixth-evidence/source-document-audit.json create mode 100644 docs/hel-892/sixth-evidence/summary.json create mode 100644 docs/hel-892/supplied-api/01.api.json create mode 100644 docs/hel-892/supplied-api/02.api.json create mode 100644 docs/hel-892/supplied-api/03.api.json create mode 100644 docs/hel-892/supplied-api/04.api.json create mode 100644 docs/hel-892/supplied-api/05.api.json create mode 100644 docs/hel-892/supplied-api/06.api.json create mode 100644 docs/hel-892/supplied-api/07.api.json create mode 100644 docs/hel-892/supplied-api/manifest.json create mode 100644 docs/hel-892/第六批真实API配套导入.md diff --git a/calliope-backend/scripts/audit_library_documents.py b/calliope-backend/scripts/audit_library_documents.py index a59203b6..92f50ac7 100644 --- a/calliope-backend/scripts/audit_library_documents.py +++ b/calliope-backend/scripts/audit_library_documents.py @@ -58,8 +58,18 @@ def audit(repo: Path, directory: Path) -> dict: visit(source) assert doc["source_links"] == source_links assert {n["node_id"]: n["source_values"] for n in doc["source_nodes"]} == nodes + source_params = [ + p for p in doc["params"] if p.get("value_origin") != "supplied_api_companion" + ] for param in doc["params"]: - assert param["default"] == nodes[param["node_id"]][param["field"]] + if param.get("value_origin") == "supplied_api_companion": + companion = doc["api_companion"] + assert companion["adopted"] + api = {n["api_id"]: n["api_node"] for n in companion["comparison"]["nodes"]} + value = api[param["api_node_id"]]["inputs"][param["field"]] + assert type(param["default"]) is type(value) and param["default"] == value + else: + assert param["default"] == nodes[param["node_id"]][param["field"]] assert param["unit"] is None assert param["min"] is None and param["max"] is None for note in doc["source_notes"]: @@ -72,7 +82,7 @@ def audit(repo: Path, directory: Path) -> dict: if fourth: assert fourth["operation_check"] in doc["steps"] assert {(f["node_id"], f["field"]) for f in checked_fields} == { - (p["node_id"], p["field"]) for p in doc["params"] + (p["node_id"], p["field"]) for p in source_params } semantic = doc["source_review"].get("semantic_review") if semantic: @@ -80,14 +90,14 @@ def audit(repo: Path, directory: Path) -> dict: assert semantic["operation_check"] in doc["steps"] assert semantic["conclusions"] and semantic["unknowns"] assert {(f["node_id"], f["field"]) for f in checked_fields} == { - (p["node_id"], p["field"]) for p in doc["params"] + (p["node_id"], p["field"]) for p in source_params } for kind in ("nodes", "notes"): for ref in semantic["evidence_refs"][kind]: key, index = ref["pointer"].strip("/").split("/") assert doc[key][int(index)]["node_id"] == ref["node_id"] assert semantic["mechanical_inventory"] == { - "named_value_count": len(doc["params"]), + "named_value_count": len(source_params), "note_count": len(doc["source_notes"]), "link_count": len(doc["source_links"]), "position_widget_node_count": len( @@ -247,12 +257,14 @@ def audit_fourth(repo: Path, report: dict, partition_path: Path) -> dict: else: assert f["class_type"] == source[nid]["class_type"] assert f["connections"]["incoming"] == [ - {"field": k, "link": v} for k, v in source[nid]["inputs"].items() + {"field": k, "link": v} + for k, v in source[nid]["inputs"].items() if isinstance(v, list) ] assert f["connections"]["outgoing"] == [ {"node_id": target, "field": k, "link": v} - for target, n in source.items() for k, v in n["inputs"].items() + for target, n in source.items() + for k, v in n["inputs"].items() if isinstance(v, list) and len(v) == 2 and v[0] == nid ] assert f["category"] in { diff --git a/calliope-backend/scripts/audit_supplied_api.py b/calliope-backend/scripts/audit_supplied_api.py new file mode 100644 index 00000000..51bbcd6b --- /dev/null +++ b/calliope-backend/scripts/audit_supplied_api.py @@ -0,0 +1,107 @@ +"""Audit received API companions against source bytes and an isolated imported DB.""" + +from __future__ import annotations + +import argparse +import hashlib +import json +from collections import Counter +from pathlib import Path + +from calliope.comfyui.companions import load_companions, resolve_companion +from calliope.comfyui.library import _json +from calliope.db import get_db + + +def audit(repo: Path, manifest: Path, db: Path, output: Path): + records = load_companions(manifest) + conn = get_db(db) + rows = [] + output.mkdir(parents=True, exist_ok=True) + try: + for path, record in records.items(): + raw = (repo / path).read_bytes() + prompt, evidence, issues = resolve_companion(record, path, raw, json.loads(raw)) + row = conn.execute( + "SELECT * FROM workflows WHERE source_repo_path=?", (path,) + ).fetchone() + assert row is not None and not row["is_enabled"] + assert row["doc_completed_at"] is row["exec_verified_at"] is None + assert row["source_sha"] == hashlib.sha1(raw).hexdigest() + assert json.loads(row["raw_source_json"]) == json.loads(raw) + assert json.loads(row["workflow_json"]) == (prompt or {}) + doc = json.loads(row["doc_json"]) + assert doc["api_companion"] == evidence + params = [p for p in doc["params"] if p.get("value_origin") == "supplied_api_companion"] + assert len(params) == sum( + not isinstance(value, list) + for node in (prompt or {}).values() + for value in node["inputs"].values() + ) + for param in params: + assert _json(param["default"]) == _json( + prompt[param["api_node_id"]]["inputs"][param["field"]] + ) + assert param["unit"] is param["min"] is param["max"] is None + details = evidence["comparison"] + fields = [f for n in details["nodes"] for f in n["fields"]] + api = json.loads(record["raw"].lstrip("\ufeff")) + assert len(fields) == sum(len(n.get("inputs", {})) for n in api.values()) + case = { + "id": evidence["binding"]["id"], + "source_path": path, + "state": row["import_state"], + "adopted": evidence["adopted"], + "api_nodes": len(api), + "api_inputs": len(fields), + "new_api_literal_params": len(params), + "field_comparison": dict(Counter(f["status"] for f in fields)), + "source_only_ids": [n["source_id"] for n in details["source_only_nodes"]], + "api_only_ids": [ + n["api_id"] for n in details["nodes"] if n["status"].startswith("api_only") + ], + "named_evidence_differences": [ + {"node_id": n["api_id"], **f} + for n in details["nodes"] + for f in n["fields"] + if f.get("named_evidence", {}).get("status") == "different" + ], + "output_map": json.loads(row["output_map"]), + "kind": row["kind"], + "issues": issues, + "general_conversion_verified": False, + "execution_verified": False, + "evidence_file": evidence["binding"]["id"] + ".json", + } + (output / case["evidence_file"]).write_text( + json.dumps(evidence, ensure_ascii=False, indent=2) + "\n", encoding="utf-8" + ) + rows.append(case) + finally: + conn.close() + report = { + "cases": rows, + "adopted": sum(r["adopted"] for r in rows), + "general_conversion_verified": 0, + "execution_verified": 0, + "new_api_literal_params": sum(r["new_api_literal_params"] for r in rows), + } + (output / "summary.json").write_text( + json.dumps(report, ensure_ascii=False, indent=2) + "\n", encoding="utf-8" + ) + return report + + +def main(): + parser = argparse.ArgumentParser(description=__doc__) + parser.add_argument("--repo", type=Path, required=True) + parser.add_argument("--manifest", type=Path, required=True) + parser.add_argument("--db", type=Path, required=True) + parser.add_argument("--output", type=Path, required=True) + args = parser.parse_args() + report = audit(args.repo, args.manifest, args.db, args.output) + print(json.dumps({k: v for k, v in report.items() if k != "cases"}, ensure_ascii=False)) + + +if __name__ == "__main__": + main() diff --git a/calliope-backend/src/calliope/comfyui/companions.py b/calliope-backend/src/calliope/comfyui/companions.py new file mode 100644 index 00000000..207c8ec3 --- /dev/null +++ b/calliope-backend/src/calliope/comfyui/companions.py @@ -0,0 +1,222 @@ +"""Explicit source-bound supplied API companions; never evidence of general conversion. + +Hashes bind the reviewed pair, not the historical export environment. Source/API +comparisons retain positional values as unnamed rather than guessing widget order. +""" + +from __future__ import annotations + +import hashlib +import json +from pathlib import Path + +from calliope.comfyui.ui2api import convert + + +def load_companions(path: Path) -> dict: + data = json.loads(path.read_text(encoding="utf-8")) + if data.get("schema_version") != 1 or not isinstance(data.get("cases"), list): + raise ValueError("配套清单格式错误") + records = {} + for entry in data["cases"]: + key = entry["source_path"] + if not all( + isinstance(entry.get(k), str) and entry[k].strip() + for k in ( + "source_path", + "source_sha", + "source_sha256", + "export_sha256", + "provider", + "attachment_id", + "received_at", + "provenance", + "instruction", + ) + ): + raise ValueError("配套清单缺少来源、哈希或说明;未知导出版本应明确记为null") + if not all(k in entry for k in ("exported_at", "frontend_version", "node_versions")): + raise ValueError("必须明确记录导出时间/版本的未知状态") + if key in records: + raise ValueError("配套源路径重复") + # Store bytes, not a live path: the same bytes are hashed and imported. + relative = Path(entry["export_file"]) + if relative.is_absolute() or ".." in relative.parts: + raise ValueError("配套文件必须在清单目录内") + raw = (path.parent / relative).read_bytes() + json.loads(raw.decode("utf-8-sig")) # Reject malformed JSON before any database write. + records[key] = {"binding": entry, "raw": raw.decode("utf-8")} + if hashlib.sha256(raw).hexdigest() != entry["export_sha256"]: + raise ValueError("配套API文件哈希不符") + return records + + +def _same(left, right): + # bool/int and int/float distinctions matter, including nested values. + return type(left) is type(right) and ( + left.keys() == right.keys() and all(_same(left[k], right[k]) for k in left) + if isinstance(left, dict) + else len(left) == len(right) and all(_same(a, b) for a, b in zip(left, right)) + if isinstance(left, list) + else left == right + ) + + +def source_comparison(source: dict, api: dict) -> dict: + """Compare every exported field with explicit source evidence, never equal-by-default.""" + definitions = {s["id"]: s for s in (source.get("definitions") or {}).get("subgraphs", [])} + indexed = {} + + def walk(graph, prefix="", ancestry=()): + for node in graph.get("nodes", []): + nid = prefix + str(node["id"]) + indexed[nid] = (node, graph, prefix) + sub = definitions.get(node.get("type")) + if sub and sub["id"] not in ancestry: + walk(sub, nid + ":", (*ancestry, sub["id"])) + + walk(source) + nodes = [] + for nid, exported in api.items(): + item = {"api_id": nid, "api_node": exported, "fields": []} + inputs = exported.get("inputs") + inputs = inputs if isinstance(inputs, dict) else {} + found = indexed.get(nid) + if not found: + item["status"] = "api_only_source_changed_or_expansion_unresolved" + item["fields"] = [ + { + "field": field, + "api_value": value, + "api_type": type(value).__name__, + "status": "source_node_unresolved", + } + for field, value in inputs.items() + ] + nodes.append(item) + continue + original, graph, prefix = found + item.update( + source_node=original, + status="type_match" + if original.get("type") == exported.get("class_type") + else "type_difference", + metadata={ + "source_title": original.get("title", original.get("type")), + "api_meta": exported.get("_meta"), + "title_equal": _same( + original.get("title", original.get("type")), + ( + exported["_meta"].get("title") + if isinstance(exported.get("_meta"), dict) + else None + ), + ), + }, + ) + links = {} + for link in graph.get("links", []): + if isinstance(link, dict): + link = [ + link.get(k) + for k in ("id", "origin_id", "origin_slot", "target_id", "target_slot", "type") + ] + if isinstance(link, list) and len(link) >= 6: + links[str(link[0])] = link + sockets = {s["name"]: s for s in original.get("inputs", []) if "name" in s} + widgets = original.get("widgets_values_named", original.get("widgets_values", [])) + for field, value in inputs.items(): + row = {"field": field, "api_value": value, "api_type": type(value).__name__} + if isinstance(widgets, dict) and field in widgets: + row["named_evidence"] = { + "source_value": widgets[field], + "source_type": type(widgets[field]).__name__, + "status": "equal" if _same(widgets[field], value) else "different", + } + socket = sockets.get(field, {}) + link = links.get(str(socket.get("link"))) + if link and len(link) >= 6: + expected = [prefix + str(link[1]), link[2]] + row.update( + source_link=link, + source_expected=expected, + status="link_equal" if _same(expected, value) else "link_difference", + ) + elif isinstance(widgets, dict) and field in widgets: + row.update( + source_value=widgets[field], + status="named_equal" if _same(widgets[field], value) else "named_difference", + ) + else: + row["status"] = "unnamed_or_virtual_boundary_unverified" + item["fields"].append(row) + nodes.append(item) + return { + "equivalent": False, + "scope": "逐节点/字段保留API值和类型;直接连线比较ID及槽;位置控件不按顺序猜名", + "nodes": nodes, + "source_only_nodes": [ + {"source_id": nid, "node": node} + for nid, (node, _, _) in indexed.items() + if nid not in api + ], + "unconfirmed": [ + "虚拟节点、子图边界、开关与循环展开语义需独立核查", + "未有可信节点定义;单位、范围、环境和实际产物未验证", + ], + } + + +def resolve_companion(record: dict, source_path: str, raw: bytes, source: dict): + binding = record["binding"] + problem = None + if ( + binding.get("source_path") != source_path + or binding.get("source_sha") != hashlib.sha1(raw).hexdigest() + or binding.get("source_sha256") != hashlib.sha256(raw).hexdigest() + ): + problem = "源文件路径或哈希变化,配套API失效" + text = record["raw"] + if hashlib.sha256(text.encode()).hexdigest() != binding.get("export_sha256"): + problem = "配套API哈希变化" + api = json.loads(text.lstrip("\ufeff")) + invalid = ( + [ + nid + for nid, node in api.items() + if not isinstance(node, dict) + or not isinstance(node.get("class_type"), str) + or not node["class_type"].strip() + or not isinstance(node.get("inputs"), dict) + or ("_meta" in node and not isinstance(node["_meta"], dict)) + ] + if isinstance(api, dict) + else ["*"] + ) + if invalid: + problem = "API缺少节点类型或inputs:" + ",".join(invalid) + checked = convert(api) + if checked.prompt is None: + problem = problem or "API连线或结构不合法" + audit = { + "binding": binding, + "adopted": not problem, + "method": "supplied_api_companion", + "general_conversion_verified": False, + "execution_verified": False, + } + if isinstance(api, dict) and all(isinstance(n, dict) for n in api.values()): + audit["comparison"] = source_comparison(source, api) + if problem: + return None, audit, [{"code": "companion_invalid", "message": problem}, *checked.issues] + # Always a blocker: no snapshot or output heuristic may promote this pair to executable. + return ( + api, + audit, + [ + { + "code": "companion_unverified", + "message": "采用提供的配套API;源图差异、导出版本与真实执行尚未核验,保持停用", + } + ], + ) diff --git a/calliope-backend/src/calliope/comfyui/library.py b/calliope-backend/src/calliope/comfyui/library.py index 47fe0c94..28a263ba 100644 --- a/calliope-backend/src/calliope/comfyui/library.py +++ b/calliope-backend/src/calliope/comfyui/library.py @@ -11,6 +11,7 @@ from datetime import datetime, timezone from pathlib import Path from typing import Any +from calliope.comfyui.companions import resolve_companion from calliope.comfyui.lint import workflow_warnings from calliope.comfyui.output_map import auto_output_map, needs_user_output_map from calliope.comfyui.parser import classify_usage, parse_dynamic_inputs, parse_dynamic_outputs @@ -300,7 +301,12 @@ def build_document( def import_library( - conn: sqlite3.Connection, root: Path, *, source_commit: str, snapshot: dict | None = None + conn: sqlite3.Connection, + root: Path, + *, + source_commit: str, + snapshot: dict | None = None, + companions: dict | None = None, ) -> dict: """Caller supplies a migrated local DB. Single transaction; no network or jobs.""" batch = str(uuid.uuid4()) @@ -319,6 +325,11 @@ def import_library( "SELECT * FROM workflows WHERE source_repo_path=?", (relative,) ).fetchone() record = {"path": relative, "source_sha": sha} + companion = (companions or {}).get(relative) + # Version companion documentation separately from the immutable received bytes. + companion_sha = hashlib.sha256( + _json({"record": companion, "doc_version": 1} if companion else None).encode() + ).hexdigest() if existing and fingerprint(dict(existing)) != existing["library_fingerprint"]: report["files"].append( { @@ -345,6 +356,8 @@ def import_library( if ( old_doc.get("snapshot_sha") == snapshot_sha and old_doc.get("source_review_sha") == review_sha + and old_doc.get("companion_sha", hashlib.sha256(b"null").hexdigest()) + == companion_sha ): report["files"].append( { @@ -357,11 +370,16 @@ def import_library( ) continue source = None + companion_audit = None try: text = raw.decode("utf-8-sig") source = json.loads(text) result = convert(source, info) prompt, issues = result.prompt, result.issues + if companion: + prompt, companion_audit, issues = resolve_companion( + companion, relative, raw, source + ) except (UnicodeDecodeError, json.JSONDecodeError) as exc: text = raw.hex() prompt, issues = ( @@ -388,6 +406,37 @@ def import_library( else "pending_test" ) doc = build_document(relative, source, prompt, snapshot, issues, review) + doc["companion_sha"] = companion_sha + if companion_audit: + doc["api_companion"] = companion_audit + if prompt: + api_doc = build_document(relative, prompt, prompt, None, [], None) + # Retain source inventory/reviews separately from supplied API evidence. + doc["params"].extend( + { + **p, + "node_id": "配套API:" + p["node_id"], + "api_node_id": p["node_id"], + "value_origin": "supplied_api_companion", + "default_source": "supplied_api_companion", + "effect": ( + f"配套API节点{p['node_id']}(" + f"{prompt[p['node_id']]['class_type']})的具名输入{p['field']};" + "现值由导出保留,含义/单位/范围仍待定义。" + ), + } + for p in api_doc["params"] + ) + doc["dependencies"]["models"] = sorted( + set(doc["dependencies"]["models"] + api_doc["dependencies"]["models"]) + ) + doc["dependencies"]["node_types"] = sorted( + set( + doc["dependencies"]["node_types"] + + api_doc["dependencies"]["node_types"] + ) + ) + doc["steps"].insert(0, companion_audit["binding"]["instruction"]) doc["snapshot_sha"] = snapshot_sha doc["source_review_sha"] = review_sha judgment = classify_usage(prompt or {}) diff --git a/calliope-backend/src/calliope/comfyui/library_cli.py b/calliope-backend/src/calliope/comfyui/library_cli.py index de30ef89..b74024ef 100644 --- a/calliope-backend/src/calliope/comfyui/library_cli.py +++ b/calliope-backend/src/calliope/comfyui/library_cli.py @@ -11,6 +11,7 @@ from pathlib import Path import httpx +from calliope.comfyui.companions import load_companions from calliope.comfyui.library import import_library, load_snapshot, rollback_batch from calliope.db import get_db, migrate_db @@ -36,6 +37,7 @@ def main() -> None: sub.add_argument("--root", type=Path, required=True) sub.add_argument("--source-commit", required=True) sub.add_argument("--snapshot", type=Path) + sub.add_argument("--companions", type=Path) else: sub.add_argument("--batch-id", required=True) args = parser.parse_args() @@ -61,12 +63,19 @@ def main() -> None: if args.command == "import" and not args.root.is_dir(): parser.error("--root 必须为存在的目录,避免误报空库成功") snapshot = load_snapshot(args.snapshot) if args.command == "import" else None + companions = ( + load_companions(args.companions) if args.command == "import" and args.companions else None + ) asyncio.run(migrate_db(args.db)) conn = get_db(args.db) try: if args.command == "import": report = import_library( - conn, args.root, source_commit=args.source_commit, snapshot=snapshot + conn, + args.root, + source_commit=args.source_commit, + snapshot=snapshot, + companions=companions, ) documents = [] for row in conn.execute( diff --git a/calliope-backend/tests/test_supplied_companions.py b/calliope-backend/tests/test_supplied_companions.py new file mode 100644 index 00000000..aa92eceb --- /dev/null +++ b/calliope-backend/tests/test_supplied_companions.py @@ -0,0 +1,229 @@ +"""Supplied real exports are disabled companions, never general conversion proof.""" + +from __future__ import annotations + +import copy +import hashlib +import json +from pathlib import Path + +import pytest + +from calliope.comfyui.companions import load_companions, resolve_companion, source_comparison +from calliope.comfyui.library import import_library, rollback_batch +from calliope.db import get_db, migrate_db + +REPO = Path(__file__).resolve().parents[2] +MANIFEST = REPO / "docs/hel-892/supplied-api/manifest.json" +ROOT = REPO / "example_ComfyUI_workflows" + + +def test_real_exports_preserved_and_case06_closed(): + records = load_companions(MANIFEST) + assert len(records) == 7 + for path, record in records.items(): + raw = (REPO / path).read_bytes() + prompt, audit, issues = resolve_companion(record, path, raw, json.loads(raw)) + assert not audit["general_conversion_verified"] + assert not audit["execution_verified"] + assert audit["binding"]["exported_at"] is None + if audit["binding"]["id"] == "06": + assert prompt is None and issues[0]["code"] == "companion_invalid" + assert "120" in issues[0]["message"] and "350" in issues[0]["message"] + else: + assert prompt == json.loads(record["raw"]) + assert audit["adopted"] and issues[0]["code"] == "companion_unverified" + # Every input, including API-only nodes, must be retained in the audit graph. + assert {n["api_id"]: n["api_node"] for n in audit["comparison"]["nodes"]} == prompt + + +@pytest.mark.parametrize("change", ["source", "path", "api", "missing_node", "negative_slot"]) +def test_mismatch_and_bad_links_fail_closed(change): + records = load_companions(MANIFEST) + path, record = next(iter(records.items())) + record = copy.deepcopy(record) + raw = (REPO / path).read_bytes() + source = json.loads(raw) + if change == "source": + raw += b" " + elif change == "path": + path += ".wrong" + elif change == "api": + record["raw"] += " " + else: + api = json.loads(record["raw"]) + api["1"]["inputs"]["image"] = ["missing", 0] if change == "missing_node" else ["2", -1] + record["raw"] = json.dumps(api) + record["binding"]["export_sha256"] = hashlib.sha256(record["raw"].encode()).hexdigest() + prompt, audit, issues = resolve_companion(record, path, raw, source) + assert prompt is None and not audit["adopted"] and issues + + +def test_slot_and_typed_named_values_and_subgraph_links(): + source = { + "nodes": [{"id": 3, "type": "group"}], + "definitions": { + "subgraphs": [ + { + "id": "group", + "nodes": [ + {"id": 1, "type": "Origin"}, + { + "id": 2, + "type": "Sink", + "widgets_values_named": {"count": 1, "seed": 42}, + "inputs": [{"name": "input", "link": 9}], + }, + ], + "links": [ + { + "id": 9, + "origin_id": 1, + "origin_slot": 2, + "target_id": 2, + "target_slot": 0, + "type": "X", + } + ], + } + ] + }, + } + api = { + "3:1": {"class_type": "Origin", "inputs": {}}, + "3:2": { + "class_type": "Sink", + "inputs": {"input": ["3:1", 1], "count": True, "seed": 43}, + "_meta": {"title": "changed"}, + }, + } + node = source_comparison(source, api)["nodes"][1] + fields = {r["field"]: r for r in node["fields"]} + assert fields["input"]["status"] == "link_difference" + assert fields["input"]["source_expected"] == ["3:1", 2] + assert fields["count"]["status"] == "named_difference" + assert fields["seed"]["named_evidence"]["status"] == "different" + assert not node["metadata"]["title_equal"] + + +async def test_pair_import_repeat_upgrade_user_protection_and_rollback(tmp_path): + await migrate_db(tmp_path / "local.db") + conn = get_db(tmp_path / "local.db") + records = load_companions(MANIFEST) + original = import_library(conn, ROOT, source_commit="fixture") + before = {r["source_repo_path"]: dict(r) for r in conn.execute("SELECT * FROM workflows")} + paired = import_library(conn, ROOT, source_commit="fixture", companions=records) + assert paired["counts"] == {"updated": 7, "skipped": 149} + assert paired["states"] == {"pending_check": 16, "needs_conversion": 140} + for path, record in records.items(): + row = conn.execute("SELECT * FROM workflows WHERE source_repo_path=?", (path,)).fetchone() + doc = json.loads(row["doc_json"]) + assert row["is_enabled"] == 0 + assert row["doc_completed_at"] is row["exec_verified_at"] is None + assert row["raw_source_json"] == before[path]["raw_source_json"] + assert doc["source_review"] == json.loads(before[path]["doc_json"])["source_review"] + if record["binding"]["id"] != "06": + assert json.loads(row["workflow_json"]) == json.loads(record["raw"]) + assert any(p["value_origin"] == "supplied_api_companion" for p in doc["params"]) + if record["binding"]["id"] == "02": + assert row["kind"] == "unknown" and json.loads(row["output_map"]) == [] + assert import_library(conn, ROOT, source_commit="fixture", companions=records)["counts"] == { + "skipped": 156 + } + assert len(rollback_batch(conn, paired["batch_id"])["files"]) == 7 + assert { + r["source_repo_path"]: dict(r) for r in conn.execute("SELECT * FROM workflows") + } == before + import_library(conn, ROOT, source_commit="fixture", companions=records) + changed = copy.deepcopy(records) + path = next(iter(changed)) + changed[path]["binding"]["instruction"] += "更新说明" + assert import_library(conn, ROOT, source_commit="fixture", companions=changed)["counts"] == { + "updated": 1, + "skipped": 155, + } + conn.execute("UPDATE workflows SET description='用户修改' WHERE source_repo_path=?", (path,)) + conn.commit() + assert import_library(conn, ROOT, source_commit="fixture", companions=records)["counts"] == { + "conflict": 1, + "skipped": 155, + } + assert ( + conn.execute( + "SELECT description FROM workflows WHERE source_repo_path=?", (path,) + ).fetchone()[0] + == "用户修改" + ) + # Removing supplied companions downgrades only managed rows, preserving the user row. + removed = import_library(conn, ROOT, source_commit="fixture") + assert removed["counts"] == {"conflict": 1, "updated": 6, "skipped": 149} + assert original["total"] == 156 + conn.close() + + +def test_file_mismatch_rejected_before_db_write(tmp_path): + data = json.loads(MANIFEST.read_text()) + case = data["cases"][0] + (tmp_path / case["export_file"]).write_text("{}") + data["cases"] = [case] + manifest = tmp_path / "manifest.json" + manifest.write_text(json.dumps(data)) + with pytest.raises(ValueError, match="哈希"): + load_companions(manifest) + + +def test_companion_routes_cannot_enable_or_test(client): + from calliope.config import settings + + conn = get_db(settings.db_path) + report = import_library( + conn, ROOT, source_commit="fixture", companions=load_companions(MANIFEST) + ) + conn.close() + case = next(r for r in report["files"] if r["path"].endswith("ControlNet_aux_常用样例.json")) + wid = case["workflow_id"] + body = client.get(f"/api/workflows/{wid}").json() + assert body["doc_json"]["api_companion"]["adopted"] + assert client.patch(f"/api/workflows/{wid}", json={"is_enabled": True}).status_code == 400 + assert client.post(f"/api/workflows/{wid}/reanalyze").json()["is_enabled"] is False + previous = settings.dry_run + settings.dry_run = False + try: + assert ( + client.post(f"/api/workflows/{wid}/test", json={"input_values": {}}).status_code == 400 + ) + finally: + settings.dry_run = previous + + +async def test_source_revision_invalidates_existing_companion(tmp_path): + root = tmp_path / "sources" + root.mkdir() + path = root / "a.json" + path.write_text(json.dumps({"nodes": [{"id": 1, "type": "Example"}]})) + raw = path.read_bytes() + api_text = json.dumps({"1": {"class_type": "Example", "inputs": {"seed": 123}}}) + record = { + "raw": api_text, + "binding": { + "source_path": "sources/a.json", + "source_sha": hashlib.sha1(raw).hexdigest(), + "source_sha256": hashlib.sha256(raw).hexdigest(), + "export_sha256": hashlib.sha256(api_text.encode()).hexdigest(), + "instruction": "夹具", + }, + } + await migrate_db(tmp_path / "local.db") + conn = get_db(tmp_path / "local.db") + records = {"sources/a.json": record} + first = import_library(conn, root, source_commit="fixture", companions=records) + assert first["states"] == {"pending_check": 1} + path.write_bytes(raw + b" ") + second = import_library(conn, root, source_commit="fixture2", companions=records) + assert second["counts"] == {"updated": 1} + assert second["states"] == {"needs_conversion": 1} + row = conn.execute("SELECT * FROM workflows").fetchone() + assert json.loads(row["workflow_json"]) == {} + assert row["raw_source_json"] == path.read_text() + assert not json.loads(row["doc_json"])["api_companion"]["adopted"] + conn.close() diff --git a/docs/hel-892/README.md b/docs/hel-892/README.md index dc1ad504..8150c97c 100644 --- a/docs/hel-892/README.md +++ b/docs/hel-892/README.md @@ -1,3 +1,5 @@ +> 当前第六批:见 `第六批真实API配套导入.md`、`supplied-api/` 和 `sixth-evidence/`。7份真实附件已收取,6份采用为停用的配套API,case06缺类型继续关闭;294项API具名现值与源侧审校分开。通用转换146份仍未证明,完整定稿/实跑仍为0,未合并部署。以下批次为历史记录。 + > 当前第四批:见 `第四批字段审校与转换前置清单.md`、`逐字段第四批审校.md`、`146份UI转换前置清单.md` 及 `fourth-evidence/`。685项新增局部审校已入实际说明;旧542项和140份说明保留。可靠转换仍剩146份,完整定稿/实跑均为0;以下为历史批次。 > 当前批次:见 `第三批语义审校与差分准备.md`、`逐文件第三批审校.md` 和 `third-evidence/`。140份源语义局部审校已补入实际说明;不要求暂停TTS或交出本地节点包/模型。下文及旧批次材料要求为历史记录,全量目标仍未完成。 diff --git a/docs/hel-892/sixth-evidence/01.json b/docs/hel-892/sixth-evidence/01.json new file mode 100644 index 00000000..9c4b7273 --- /dev/null +++ b/docs/hel-892/sixth-evidence/01.json @@ -0,0 +1,2876 @@ +{ + "binding": { + "id": "01", + "source_path": "example_ComfyUI_workflows/0-基础处理:图像合并-音频分离-视频截取合并-图像视频放大/ControlNet_aux_常用样例.json", + "source_sha": "03d2fe6ae190c7cae17b89947f8975016616014a", + "export_file": "01.api.json", + "source_sha256": "8898689d777330c4749a079cd762f573bb12a278c47a2fefbd066ffbdc122b6b", + "export_sha256": "0ed76e088a23af43d6fecbf273fc0a983d7112e2b046d347f715afffac56d57e", + "original_filename": "ControlNet_aux_常用样例.json", + "provider": "白栖知", + "attachment_id": "01a187e2-366f-7cdc-88cf-e0ee34746154", + "received_at": "2026-10-29T13:38:38Z", + "received_at_basis": "主卡附件评论创建时间;不是导出时间", + "source_comment_id": "01a187e2-4f6e-761e-9412-43042e094daa", + "exported_at": null, + "frontend_version": null, + "node_versions": null, + "provenance": "user_supplied_api_export_environment_unknown", + "source_export_equivalence_verified": false, + "instruction": "图像预处理:API节点2和38各读一张图;1/4/10/12/18/19/20/21/33/34按各自具名参数预处理,28/29/30拼接,多个PreviewImage仅作预览。检查两张输入文件及分支连线,不能把Reroute剔除当通用转换规则。" + }, + "adopted": true, + "method": "supplied_api_companion", + "general_conversion_verified": false, + "execution_verified": false, + "comparison": { + "equivalent": false, + "scope": "逐节点/字段保留API值和类型;直接连线比较ID及槽;位置控件不按顺序猜名", + "nodes": [ + { + "api_id": "1", + "api_node": { + "inputs": { + "preprocessor": "DepthAnythingPreprocessor", + "resolution": 512, + "image": [ + "2", + 0 + ] + }, + "class_type": "AIO_Preprocessor", + "_meta": { + "title": "Aux集成预处理器" + } + }, + "fields": [ + { + "field": "preprocessor", + "api_value": "DepthAnythingPreprocessor", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "resolution", + "api_value": 512, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "image", + "api_value": [ + "2", + 0 + ], + "api_type": "list", + "source_link": [ + 46, + 32, + 0, + 1, + 0, + "IMAGE" + ], + "source_expected": [ + "32", + 0 + ], + "status": "link_difference" + } + ], + "source_node": { + "id": 1, + "type": "AIO_Preprocessor", + "pos": [ + 419.2714440229391, + -1339.0887265051863 + ], + "size": [ + 270, + 82 + ], + "flags": {}, + "order": 14, + "mode": 0, + "inputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "image", + "type": "IMAGE", + "link": 46 + }, + { + "label": "预处理器", + "localized_name": "预处理器", + "name": "preprocessor", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "preprocessor" + }, + "link": null + }, + { + "label": "分辨率", + "localized_name": "分辨率", + "name": "resolution", + "shape": 7, + "type": "INT", + "widget": { + "name": "resolution" + }, + "link": null + } + ], + "outputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 2 + ] + } + ], + "title": "Aux集成预处理器", + "properties": { + "cnr_id": "comfyui_controlnet_aux", + "ver": "1.1.5", + "Node name for S&R": "AIO_Preprocessor", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [ + "DepthAnythingPreprocessor", + 512 + ] + }, + "status": "type_match", + "metadata": { + "source_title": "Aux集成预处理器", + "api_meta": { + "title": "Aux集成预处理器" + }, + "title_equal": true + } + }, + { + "api_id": "2", + "api_node": { + "inputs": { + "image": "comfyui-airport-editorial_00001_.png" + }, + "class_type": "LoadImage", + "_meta": { + "title": "加载图像" + } + }, + "fields": [ + { + "field": "image", + "api_value": "comfyui-airport-editorial_00001_.png", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + } + ], + "source_node": { + "id": 2, + "type": "LoadImage", + "pos": [ + -482.2702471404147, + -593.1848228908528 + ], + "size": [ + 270, + 314 + ], + "flags": {}, + "order": 0, + "mode": 0, + "inputs": [ + { + "localized_name": "图像", + "name": "image", + "type": "COMBO", + "widget": { + "name": "image" + }, + "link": null + }, + { + "localized_name": "选择文件上传", + "name": "upload", + "type": "IMAGEUPLOAD", + "widget": { + "name": "upload" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 22, + 41, + 45, + 47 + ] + }, + { + "localized_name": "遮罩", + "name": "MASK", + "type": "MASK", + "links": null + } + ], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.21.1", + "Node name for S&R": "LoadImage", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [ + "comfyui-airport-editorial_00001_.png", + "image" + ] + }, + "status": "type_match", + "metadata": { + "source_title": "LoadImage", + "api_meta": { + "title": "加载图像" + }, + "title_equal": false + } + }, + { + "api_id": "3", + "api_node": { + "inputs": { + "images": [ + "1", + 0 + ] + }, + "class_type": "PreviewImage", + "_meta": { + "title": "预览图像" + } + }, + "fields": [ + { + "field": "images", + "api_value": [ + "1", + 0 + ], + "api_type": "list", + "source_link": [ + 2, + 1, + 0, + 3, + 0, + "IMAGE" + ], + "source_expected": [ + "1", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 3, + "type": "PreviewImage", + "pos": [ + 994.1136629847879, + -1504.1025926771606 + ], + "size": [ + 210, + 276.066120618284 + ], + "flags": {}, + "order": 22, + "mode": 0, + "inputs": [ + { + "localized_name": "图像", + "name": "images", + "type": "IMAGE", + "link": 2 + } + ], + "outputs": [], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.21.1", + "Node name for S&R": "PreviewImage", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [] + }, + "status": "type_match", + "metadata": { + "source_title": "PreviewImage", + "api_meta": { + "title": "预览图像" + }, + "title_equal": false + } + }, + { + "api_id": "4", + "api_node": { + "inputs": { + "ckpt_name": "depth_anything_vitl14.pth", + "resolution": 512, + "image": [ + "2", + 0 + ] + }, + "class_type": "DepthAnythingPreprocessor", + "_meta": { + "title": "DepthAnything任何深度预处理器" + } + }, + "fields": [ + { + "field": "ckpt_name", + "api_value": "depth_anything_vitl14.pth", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "resolution", + "api_value": 512, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "image", + "api_value": [ + "2", + 0 + ], + "api_type": "list", + "source_link": [ + 44, + 32, + 0, + 4, + 0, + "IMAGE" + ], + "source_expected": [ + "32", + 0 + ], + "status": "link_difference" + } + ], + "source_node": { + "id": 4, + "type": "DepthAnythingPreprocessor", + "pos": [ + 416.53876886400724, + -1511.838806993518 + ], + "size": [ + 272.74410870029453, + 82 + ], + "flags": {}, + "order": 13, + "mode": 0, + "inputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "image", + "type": "IMAGE", + "link": 44 + }, + { + "label": "模型名称", + "localized_name": "模型名称", + "name": "ckpt_name", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "ckpt_name" + }, + "link": null + }, + { + "label": "分辨率", + "localized_name": "分辨率", + "name": "resolution", + "shape": 7, + "type": "INT", + "widget": { + "name": "resolution" + }, + "link": null + } + ], + "outputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 4 + ] + } + ], + "title": "DepthAnything任何深度预处理器", + "properties": { + "cnr_id": "comfyui_controlnet_aux", + "ver": "1.1.5", + "Node name for S&R": "DepthAnythingPreprocessor", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [ + "depth_anything_vitl14.pth", + 512 + ] + }, + "status": "type_match", + "metadata": { + "source_title": "DepthAnything任何深度预处理器", + "api_meta": { + "title": "DepthAnything任何深度预处理器" + }, + "title_equal": true + } + }, + { + "api_id": "5", + "api_node": { + "inputs": { + "images": [ + "4", + 0 + ] + }, + "class_type": "PreviewImage", + "_meta": { + "title": "预览图像" + } + }, + "fields": [ + { + "field": "images", + "api_value": [ + "4", + 0 + ], + "api_type": "list", + "source_link": [ + 4, + 4, + 0, + 5, + 0, + "IMAGE" + ], + "source_expected": [ + "4", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 5, + "type": "PreviewImage", + "pos": [ + 743.096516427487, + -1505.756928523272 + ], + "size": [ + 210, + 299.83933226995055 + ], + "flags": {}, + "order": 21, + "mode": 0, + "inputs": [ + { + "localized_name": "图像", + "name": "images", + "type": "IMAGE", + "link": 4 + } + ], + "outputs": [], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.21.1", + "Node name for S&R": "PreviewImage", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [] + }, + "status": "type_match", + "metadata": { + "source_title": "PreviewImage", + "api_meta": { + "title": "预览图像" + }, + "title_equal": false + } + }, + { + "api_id": "10", + "api_node": { + "inputs": { + "low_threshold": 0.4, + "high_threshold": 0.8, + "image": [ + "2", + 0 + ] + }, + "class_type": "Canny", + "_meta": { + "title": "Canny边缘检测" + } + }, + "fields": [ + { + "field": "low_threshold", + "api_value": 0.4, + "api_type": "float", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "high_threshold", + "api_value": 0.8, + "api_type": "float", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "image", + "api_value": [ + "2", + 0 + ], + "api_type": "list", + "source_link": [ + 42, + 31, + 0, + 10, + 0, + "IMAGE" + ], + "source_expected": [ + "31", + 0 + ], + "status": "link_difference" + } + ], + "source_node": { + "id": 10, + "type": "Canny", + "pos": [ + 469.5024025876337, + -1033.5905515977925 + ], + "size": [ + 270, + 82 + ], + "flags": {}, + "order": 11, + "mode": 0, + "inputs": [ + { + "localized_name": "图像", + "name": "image", + "type": "IMAGE", + "link": 42 + }, + { + "localized_name": "低阈值", + "name": "low_threshold", + "type": "FLOAT", + "widget": { + "name": "low_threshold" + }, + "link": null + }, + { + "localized_name": "高阈值", + "name": "high_threshold", + "type": "FLOAT", + "widget": { + "name": "high_threshold" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 8 + ] + } + ], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.21.1", + "Node name for S&R": "Canny", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [ + 0.4, + 0.8 + ] + }, + "status": "type_match", + "metadata": { + "source_title": "Canny", + "api_meta": { + "title": "Canny边缘检测" + }, + "title_equal": false + } + }, + { + "api_id": "11", + "api_node": { + "inputs": { + "images": [ + "10", + 0 + ] + }, + "class_type": "PreviewImage", + "_meta": { + "title": "预览图像" + } + }, + "fields": [ + { + "field": "images", + "api_value": [ + "10", + 0 + ], + "api_type": "list", + "source_link": [ + 8, + 10, + 0, + 11, + 0, + "IMAGE" + ], + "source_expected": [ + "10", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 11, + "type": "PreviewImage", + "pos": [ + 791.5967586530584, + -1011.7974527994733 + ], + "size": [ + 276.90573062930275, + 285.2257992252497 + ], + "flags": {}, + "order": 19, + "mode": 0, + "inputs": [ + { + "localized_name": "图像", + "name": "images", + "type": "IMAGE", + "link": 8 + } + ], + "outputs": [], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.21.1", + "Node name for S&R": "PreviewImage", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [] + }, + "status": "type_match", + "metadata": { + "source_title": "PreviewImage", + "api_meta": { + "title": "预览图像" + }, + "title_equal": false + } + }, + { + "api_id": "12", + "api_node": { + "inputs": { + "preprocessor": "CannyEdgePreprocessor", + "resolution": 1024, + "image": [ + "2", + 0 + ] + }, + "class_type": "AIO_Preprocessor", + "_meta": { + "title": "Aux集成预处理器" + } + }, + "fields": [ + { + "field": "preprocessor", + "api_value": "CannyEdgePreprocessor", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "resolution", + "api_value": 1024, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "image", + "api_value": [ + "2", + 0 + ], + "api_type": "list", + "source_link": [ + 43, + 31, + 0, + 12, + 0, + "IMAGE" + ], + "source_expected": [ + "31", + 0 + ], + "status": "link_difference" + } + ], + "source_node": { + "id": 12, + "type": "AIO_Preprocessor", + "pos": [ + 472.08362302315527, + -853.4564144778592 + ], + "size": [ + 270, + 82 + ], + "flags": {}, + "order": 12, + "mode": 0, + "inputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "image", + "type": "IMAGE", + "link": 43 + }, + { + "label": "预处理器", + "localized_name": "预处理器", + "name": "preprocessor", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "preprocessor" + }, + "link": null + }, + { + "label": "分辨率", + "localized_name": "分辨率", + "name": "resolution", + "shape": 7, + "type": "INT", + "widget": { + "name": "resolution" + }, + "link": null + } + ], + "outputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 10 + ] + } + ], + "title": "Aux集成预处理器", + "properties": { + "cnr_id": "comfyui_controlnet_aux", + "ver": "1.1.5", + "Node name for S&R": "AIO_Preprocessor", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [ + "CannyEdgePreprocessor", + 1024 + ] + }, + "status": "type_match", + "metadata": { + "source_title": "Aux集成预处理器", + "api_meta": { + "title": "Aux集成预处理器" + }, + "title_equal": true + } + }, + { + "api_id": "13", + "api_node": { + "inputs": { + "images": [ + "12", + 0 + ] + }, + "class_type": "PreviewImage", + "_meta": { + "title": "预览图像" + } + }, + "fields": [ + { + "field": "images", + "api_value": [ + "12", + 0 + ], + "api_type": "list", + "source_link": [ + 10, + 12, + 0, + 13, + 0, + "IMAGE" + ], + "source_expected": [ + "12", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 13, + "type": "PreviewImage", + "pos": [ + 1094.1547770393574, + -1010.1511146107904 + ], + "size": [ + 259.21566431203314, + 293.65839471346817 + ], + "flags": {}, + "order": 20, + "mode": 0, + "inputs": [ + { + "localized_name": "图像", + "name": "images", + "type": "IMAGE", + "link": 10 + } + ], + "outputs": [], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.21.1", + "Node name for S&R": "PreviewImage", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [] + }, + "status": "type_match", + "metadata": { + "source_title": "PreviewImage", + "api_meta": { + "title": "预览图像" + }, + "title_equal": false + } + }, + { + "api_id": "18", + "api_node": { + "inputs": { + "preprocessor": "AnyLineArtPreprocessor_aux", + "resolution": 1024, + "image": [ + "2", + 0 + ] + }, + "class_type": "AIO_Preprocessor", + "_meta": { + "title": "Aux集成预处理器" + } + }, + "fields": [ + { + "field": "preprocessor", + "api_value": "AnyLineArtPreprocessor_aux", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "resolution", + "api_value": 1024, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "image", + "api_value": [ + "2", + 0 + ], + "api_type": "list", + "source_link": [ + 23, + 24, + 0, + 18, + 0, + "IMAGE" + ], + "source_expected": [ + "24", + 0 + ], + "status": "link_difference" + } + ], + "source_node": { + "id": 18, + "type": "AIO_Preprocessor", + "pos": [ + 454.4392129313587, + -601.7920180215307 + ], + "size": [ + 270, + 82 + ], + "flags": {}, + "order": 7, + "mode": 0, + "inputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "image", + "type": "IMAGE", + "link": 23 + }, + { + "label": "预处理器", + "localized_name": "预处理器", + "name": "preprocessor", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "preprocessor" + }, + "link": null + }, + { + "label": "分辨率", + "localized_name": "分辨率", + "name": "resolution", + "shape": 7, + "type": "INT", + "widget": { + "name": "resolution" + }, + "link": null + } + ], + "outputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 35 + ] + } + ], + "title": "Aux集成预处理器", + "properties": { + "cnr_id": "comfyui_controlnet_aux", + "ver": "1.1.5", + "Node name for S&R": "AIO_Preprocessor", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [ + "AnyLineArtPreprocessor_aux", + 1024 + ] + }, + "status": "type_match", + "metadata": { + "source_title": "Aux集成预处理器", + "api_meta": { + "title": "Aux集成预处理器" + }, + "title_equal": true + } + }, + { + "api_id": "19", + "api_node": { + "inputs": { + "preprocessor": "LineArtPreprocessor", + "resolution": 1024, + "image": [ + "2", + 0 + ] + }, + "class_type": "AIO_Preprocessor", + "_meta": { + "title": "Aux集成预处理器" + } + }, + "fields": [ + { + "field": "preprocessor", + "api_value": "LineArtPreprocessor", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "resolution", + "api_value": 1024, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "image", + "api_value": [ + "2", + 0 + ], + "api_type": "list", + "source_link": [ + 24, + 24, + 0, + 19, + 0, + "IMAGE" + ], + "source_expected": [ + "24", + 0 + ], + "status": "link_difference" + } + ], + "source_node": { + "id": 19, + "type": "AIO_Preprocessor", + "pos": [ + 450.6022584614121, + -468.3591423696305 + ], + "size": [ + 270, + 82 + ], + "flags": {}, + "order": 8, + "mode": 0, + "inputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "image", + "type": "IMAGE", + "link": 24 + }, + { + "label": "预处理器", + "localized_name": "预处理器", + "name": "preprocessor", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "preprocessor" + }, + "link": null + }, + { + "label": "分辨率", + "localized_name": "分辨率", + "name": "resolution", + "shape": 7, + "type": "INT", + "widget": { + "name": "resolution" + }, + "link": null + } + ], + "outputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 36 + ] + } + ], + "title": "Aux集成预处理器", + "properties": { + "cnr_id": "comfyui_controlnet_aux", + "ver": "1.1.5", + "Node name for S&R": "AIO_Preprocessor", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [ + "LineArtPreprocessor", + 1024 + ] + }, + "status": "type_match", + "metadata": { + "source_title": "Aux集成预处理器", + "api_meta": { + "title": "Aux集成预处理器" + }, + "title_equal": true + } + }, + { + "api_id": "20", + "api_node": { + "inputs": { + "preprocessor": "AnimeLineArtPreprocessor", + "resolution": 1024, + "image": [ + "2", + 0 + ] + }, + "class_type": "AIO_Preprocessor", + "_meta": { + "title": "Aux集成预处理器" + } + }, + "fields": [ + { + "field": "preprocessor", + "api_value": "AnimeLineArtPreprocessor", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "resolution", + "api_value": 1024, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "image", + "api_value": [ + "2", + 0 + ], + "api_type": "list", + "source_link": [ + 25, + 24, + 0, + 20, + 0, + "IMAGE" + ], + "source_expected": [ + "24", + 0 + ], + "status": "link_difference" + } + ], + "source_node": { + "id": 20, + "type": "AIO_Preprocessor", + "pos": [ + 452.000682676216, + -335.50884196325933 + ], + "size": [ + 270, + 82 + ], + "flags": {}, + "order": 9, + "mode": 0, + "inputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "image", + "type": "IMAGE", + "link": 25 + }, + { + "label": "预处理器", + "localized_name": "预处理器", + "name": "preprocessor", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "preprocessor" + }, + "link": null + }, + { + "label": "分辨率", + "localized_name": "分辨率", + "name": "resolution", + "shape": 7, + "type": "INT", + "widget": { + "name": "resolution" + }, + "link": null + } + ], + "outputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 37 + ] + } + ], + "title": "Aux集成预处理器", + "properties": { + "cnr_id": "comfyui_controlnet_aux", + "ver": "1.1.5", + "Node name for S&R": "AIO_Preprocessor", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [ + "AnimeLineArtPreprocessor", + 1024 + ] + }, + "status": "type_match", + "metadata": { + "source_title": "Aux集成预处理器", + "api_meta": { + "title": "Aux集成预处理器" + }, + "title_equal": true + } + }, + { + "api_id": "21", + "api_node": { + "inputs": { + "preprocessor": "LineartStandardPreprocessor", + "resolution": 1024, + "image": [ + "2", + 0 + ] + }, + "class_type": "AIO_Preprocessor", + "_meta": { + "title": "Aux集成预处理器" + } + }, + "fields": [ + { + "field": "preprocessor", + "api_value": "LineartStandardPreprocessor", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "resolution", + "api_value": 1024, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "image", + "api_value": [ + "2", + 0 + ], + "api_type": "list", + "source_link": [ + 26, + 24, + 0, + 21, + 0, + "IMAGE" + ], + "source_expected": [ + "24", + 0 + ], + "status": "link_difference" + } + ], + "source_node": { + "id": 21, + "type": "AIO_Preprocessor", + "pos": [ + 452.00068267621594, + -184.05530422735177 + ], + "size": [ + 270, + 82 + ], + "flags": {}, + "order": 10, + "mode": 0, + "inputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "image", + "type": "IMAGE", + "link": 26 + }, + { + "label": "预处理器", + "localized_name": "预处理器", + "name": "preprocessor", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "preprocessor" + }, + "link": null + }, + { + "label": "分辨率", + "localized_name": "分辨率", + "name": "resolution", + "shape": 7, + "type": "INT", + "widget": { + "name": "resolution" + }, + "link": null + } + ], + "outputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 38 + ] + } + ], + "title": "Aux集成预处理器", + "properties": { + "cnr_id": "comfyui_controlnet_aux", + "ver": "1.1.5", + "Node name for S&R": "AIO_Preprocessor", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [ + "LineartStandardPreprocessor", + 1024 + ] + }, + "status": "type_match", + "metadata": { + "source_title": "Aux集成预处理器", + "api_meta": { + "title": "Aux集成预处理器" + }, + "title_equal": true + } + }, + { + "api_id": "22", + "api_node": { + "inputs": { + "images": [ + "30", + 0 + ] + }, + "class_type": "PreviewImage", + "_meta": { + "title": "预览图像" + } + }, + "fields": [ + { + "field": "images", + "api_value": [ + "30", + 0 + ], + "api_type": "list", + "source_link": [ + 39, + 30, + 0, + 22, + 0, + "IMAGE" + ], + "source_expected": [ + "30", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 22, + "type": "PreviewImage", + "pos": [ + 1388.6214359360067, + -583.7500095726137 + ], + "size": [ + 967.2559538608341, + 586.1530556149563 + ], + "flags": {}, + "order": 25, + "mode": 0, + "inputs": [ + { + "localized_name": "图像", + "name": "images", + "type": "IMAGE", + "link": 39 + } + ], + "outputs": [], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.21.1", + "Node name for S&R": "PreviewImage", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [] + }, + "status": "type_match", + "metadata": { + "source_title": "PreviewImage", + "api_meta": { + "title": "预览图像" + }, + "title_equal": false + } + }, + { + "api_id": "28", + "api_node": { + "inputs": { + "direction": "right", + "match_image_size": true, + "spacing_width": 2, + "spacing_color": "white", + "image1": [ + "18", + 0 + ], + "image2": [ + "19", + 0 + ] + }, + "class_type": "ImageStitch", + "_meta": { + "title": "图像拼接" + } + }, + "fields": [ + { + "field": "direction", + "api_value": "right", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "match_image_size", + "api_value": true, + "api_type": "bool", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "spacing_width", + "api_value": 2, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "spacing_color", + "api_value": "white", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "image1", + "api_value": [ + "18", + 0 + ], + "api_type": "list", + "source_link": [ + 35, + 18, + 0, + 28, + 0, + "IMAGE" + ], + "source_expected": [ + "18", + 0 + ], + "status": "link_equal" + }, + { + "field": "image2", + "api_value": [ + "19", + 0 + ], + "api_type": "list", + "source_link": [ + 36, + 19, + 0, + 28, + 1, + "IMAGE" + ], + "source_expected": [ + "19", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 28, + "type": "ImageStitch", + "pos": [ + 830.2586130368217, + -591.3246364640868 + ], + "size": [ + 230, + 160 + ], + "flags": {}, + "order": 17, + "mode": 0, + "inputs": [ + { + "label": "图像1", + "localized_name": "图像1", + "name": "image1", + "type": "IMAGE", + "link": 35 + }, + { + "label": "图像2", + "localized_name": "图像2", + "name": "image2", + "shape": 7, + "type": "IMAGE", + "link": 36 + }, + { + "label": "方向", + "localized_name": "方向", + "name": "direction", + "type": "COMBO", + "widget": { + "name": "direction" + }, + "link": null + }, + { + "label": "匹配图像尺寸", + "localized_name": "匹配图像尺寸", + "name": "match_image_size", + "type": "BOOLEAN", + "widget": { + "name": "match_image_size" + }, + "link": null + }, + { + "label": "间距宽度", + "localized_name": "间距宽度", + "name": "spacing_width", + "type": "INT", + "widget": { + "name": "spacing_width" + }, + "link": null + }, + { + "label": "间距颜色", + "localized_name": "间距颜色", + "name": "spacing_color", + "type": "COMBO", + "widget": { + "name": "spacing_color" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 33 + ] + } + ], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.17.0", + "Node name for S&R": "ImageStitch", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.7" + } + }, + "widgets_values": [ + "right", + true, + 2, + "white" + ] + }, + "status": "type_match", + "metadata": { + "source_title": "ImageStitch", + "api_meta": { + "title": "图像拼接" + }, + "title_equal": false + } + }, + { + "api_id": "29", + "api_node": { + "inputs": { + "direction": "right", + "match_image_size": true, + "spacing_width": 2, + "spacing_color": "white", + "image1": [ + "20", + 0 + ], + "image2": [ + "21", + 0 + ] + }, + "class_type": "ImageStitch", + "_meta": { + "title": "图像拼接" + } + }, + "fields": [ + { + "field": "direction", + "api_value": "right", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "match_image_size", + "api_value": true, + "api_type": "bool", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "spacing_width", + "api_value": 2, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "spacing_color", + "api_value": "white", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "image1", + "api_value": [ + "20", + 0 + ], + "api_type": "list", + "source_link": [ + 37, + 20, + 0, + 29, + 0, + "IMAGE" + ], + "source_expected": [ + "20", + 0 + ], + "status": "link_equal" + }, + { + "field": "image2", + "api_value": [ + "21", + 0 + ], + "api_type": "list", + "source_link": [ + 38, + 21, + 0, + 29, + 1, + "IMAGE" + ], + "source_expected": [ + "21", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 29, + "type": "ImageStitch", + "pos": [ + 828.2949387622735, + -370.90554013067674 + ], + "size": [ + 230, + 160 + ], + "flags": {}, + "order": 18, + "mode": 0, + "inputs": [ + { + "label": "图像1", + "localized_name": "图像1", + "name": "image1", + "type": "IMAGE", + "link": 37 + }, + { + "label": "图像2", + "localized_name": "图像2", + "name": "image2", + "shape": 7, + "type": "IMAGE", + "link": 38 + }, + { + "label": "方向", + "localized_name": "方向", + "name": "direction", + "type": "COMBO", + "widget": { + "name": "direction" + }, + "link": null + }, + { + "label": "匹配图像尺寸", + "localized_name": "匹配图像尺寸", + "name": "match_image_size", + "type": "BOOLEAN", + "widget": { + "name": "match_image_size" + }, + "link": null + }, + { + "label": "间距宽度", + "localized_name": "间距宽度", + "name": "spacing_width", + "type": "INT", + "widget": { + "name": "spacing_width" + }, + "link": null + }, + { + "label": "间距颜色", + "localized_name": "间距颜色", + "name": "spacing_color", + "type": "COMBO", + "widget": { + "name": "spacing_color" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 34 + ] + } + ], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.17.0", + "Node name for S&R": "ImageStitch", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.7" + } + }, + "widgets_values": [ + "right", + true, + 2, + "white" + ] + }, + "status": "type_match", + "metadata": { + "source_title": "ImageStitch", + "api_meta": { + "title": "图像拼接" + }, + "title_equal": false + } + }, + { + "api_id": "30", + "api_node": { + "inputs": { + "direction": "down", + "match_image_size": true, + "spacing_width": 2, + "spacing_color": "white", + "image1": [ + "28", + 0 + ], + "image2": [ + "29", + 0 + ] + }, + "class_type": "ImageStitch", + "_meta": { + "title": "图像拼接" + } + }, + "fields": [ + { + "field": "direction", + "api_value": "down", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "match_image_size", + "api_value": true, + "api_type": "bool", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "spacing_width", + "api_value": 2, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "spacing_color", + "api_value": "white", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "image1", + "api_value": [ + "28", + 0 + ], + "api_type": "list", + "source_link": [ + 33, + 28, + 0, + 30, + 0, + "IMAGE" + ], + "source_expected": [ + "28", + 0 + ], + "status": "link_equal" + }, + { + "field": "image2", + "api_value": [ + "29", + 0 + ], + "api_type": "list", + "source_link": [ + 34, + 29, + 0, + 30, + 1, + "IMAGE" + ], + "source_expected": [ + "29", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 30, + "type": "ImageStitch", + "pos": [ + 1104.2183827611025, + -582.9487746295188 + ], + "size": [ + 230, + 160 + ], + "flags": {}, + "order": 24, + "mode": 0, + "showAdvanced": false, + "inputs": [ + { + "label": "图像1", + "localized_name": "图像1", + "name": "image1", + "type": "IMAGE", + "link": 33 + }, + { + "label": "图像2", + "localized_name": "图像2", + "name": "image2", + "shape": 7, + "type": "IMAGE", + "link": 34 + }, + { + "label": "方向", + "localized_name": "方向", + "name": "direction", + "type": "COMBO", + "widget": { + "name": "direction" + }, + "link": null + }, + { + "label": "匹配图像尺寸", + "localized_name": "匹配图像尺寸", + "name": "match_image_size", + "type": "BOOLEAN", + "widget": { + "name": "match_image_size" + }, + "link": null + }, + { + "label": "间距宽度", + "localized_name": "间距宽度", + "name": "spacing_width", + "type": "INT", + "widget": { + "name": "spacing_width" + }, + "link": null + }, + { + "label": "间距颜色", + "localized_name": "间距颜色", + "name": "spacing_color", + "type": "COMBO", + "widget": { + "name": "spacing_color" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 39 + ] + } + ], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.17.0", + "Node name for S&R": "ImageStitch", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.7" + } + }, + "widgets_values": [ + "down", + true, + 2, + "white" + ] + }, + "status": "type_match", + "metadata": { + "source_title": "ImageStitch", + "api_meta": { + "title": "图像拼接" + }, + "title_equal": false + } + }, + { + "api_id": "33", + "api_node": { + "inputs": { + "preprocessor": "DWPreprocessor", + "resolution": 1024, + "image": [ + "2", + 0 + ] + }, + "class_type": "AIO_Preprocessor", + "_meta": { + "title": "Aux集成预处理器" + } + }, + "fields": [ + { + "field": "preprocessor", + "api_value": "DWPreprocessor", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "resolution", + "api_value": 1024, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "image", + "api_value": [ + "2", + 0 + ], + "api_type": "list", + "source_link": [ + 48, + 35, + 0, + 33, + 0, + "IMAGE" + ], + "source_expected": [ + "35", + 0 + ], + "status": "link_difference" + } + ], + "source_node": { + "id": 33, + "type": "AIO_Preprocessor", + "pos": [ + 422.4888280899862, + 34.74703769927764 + ], + "size": [ + 270, + 82 + ], + "flags": {}, + "order": 15, + "mode": 0, + "inputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "image", + "type": "IMAGE", + "link": 48 + }, + { + "label": "预处理器", + "localized_name": "预处理器", + "name": "preprocessor", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "preprocessor" + }, + "link": null + }, + { + "label": "分辨率", + "localized_name": "分辨率", + "name": "resolution", + "shape": 7, + "type": "INT", + "widget": { + "name": "resolution" + }, + "link": null + } + ], + "outputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 49 + ] + } + ], + "title": "Aux集成预处理器", + "properties": { + "cnr_id": "comfyui_controlnet_aux", + "ver": "1.1.5", + "Node name for S&R": "AIO_Preprocessor", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [ + "DWPreprocessor", + 1024 + ] + }, + "status": "type_match", + "metadata": { + "source_title": "Aux集成预处理器", + "api_meta": { + "title": "Aux集成预处理器" + }, + "title_equal": true + } + }, + { + "api_id": "34", + "api_node": { + "inputs": { + "preprocessor": "AnimalPosePreprocessor", + "resolution": 1024, + "image": [ + "38", + 0 + ] + }, + "class_type": "AIO_Preprocessor", + "_meta": { + "title": "Aux集成预处理器" + } + }, + "fields": [ + { + "field": "preprocessor", + "api_value": "AnimalPosePreprocessor", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "resolution", + "api_value": 1024, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "image", + "api_value": [ + "38", + 0 + ], + "api_type": "list", + "source_link": [ + 52, + 38, + 0, + 34, + 0, + "IMAGE" + ], + "source_expected": [ + "38", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 34, + "type": "AIO_Preprocessor", + "pos": [ + 426.9730634656538, + 306.1884396426034 + ], + "size": [ + 270, + 82 + ], + "flags": {}, + "order": 6, + "mode": 0, + "inputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "image", + "type": "IMAGE", + "link": 52 + }, + { + "label": "预处理器", + "localized_name": "预处理器", + "name": "preprocessor", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "preprocessor" + }, + "link": null + }, + { + "label": "分辨率", + "localized_name": "分辨率", + "name": "resolution", + "shape": 7, + "type": "INT", + "widget": { + "name": "resolution" + }, + "link": null + } + ], + "outputs": [ + { + "label": "图像", + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 51 + ] + } + ], + "title": "Aux集成预处理器", + "properties": { + "cnr_id": "comfyui_controlnet_aux", + "ver": "1.1.5", + "Node name for S&R": "AIO_Preprocessor", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [ + "AnimalPosePreprocessor", + 1024 + ] + }, + "status": "type_match", + "metadata": { + "source_title": "Aux集成预处理器", + "api_meta": { + "title": "Aux集成预处理器" + }, + "title_equal": true + } + }, + { + "api_id": "36", + "api_node": { + "inputs": { + "images": [ + "33", + 0 + ] + }, + "class_type": "PreviewImage", + "_meta": { + "title": "预览图像" + } + }, + "fields": [ + { + "field": "images", + "api_value": [ + "33", + 0 + ], + "api_type": "list", + "source_link": [ + 49, + 33, + 0, + 36, + 0, + "IMAGE" + ], + "source_expected": [ + "33", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 36, + "type": "PreviewImage", + "pos": [ + 928.6235375696226, + -3.9155061676062997 + ], + "size": [ + 362.14646887904246, + 295.1383894294655 + ], + "flags": {}, + "order": 23, + "mode": 0, + "inputs": [ + { + "localized_name": "图像", + "name": "images", + "type": "IMAGE", + "link": 49 + } + ], + "outputs": [], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.21.1", + "Node name for S&R": "PreviewImage", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [] + }, + "status": "type_match", + "metadata": { + "source_title": "PreviewImage", + "api_meta": { + "title": "预览图像" + }, + "title_equal": false + } + }, + { + "api_id": "37", + "api_node": { + "inputs": { + "images": [ + "34", + 0 + ] + }, + "class_type": "PreviewImage", + "_meta": { + "title": "预览图像" + } + }, + "fields": [ + { + "field": "images", + "api_value": [ + "34", + 0 + ], + "api_type": "list", + "source_link": [ + 51, + 34, + 0, + 37, + 0, + "IMAGE" + ], + "source_expected": [ + "34", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 37, + "type": "PreviewImage", + "pos": [ + 942.8997145769215, + 341.9656216393168 + ], + "size": [ + 346.43538258935223, + 258 + ], + "flags": {}, + "order": 16, + "mode": 0, + "inputs": [ + { + "localized_name": "图像", + "name": "images", + "type": "IMAGE", + "link": 51 + } + ], + "outputs": [], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.21.1", + "Node name for S&R": "PreviewImage", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [] + }, + "status": "type_match", + "metadata": { + "source_title": "PreviewImage", + "api_meta": { + "title": "预览图像" + }, + "title_equal": false + } + }, + { + "api_id": "38", + "api_node": { + "inputs": { + "image": "1772086821573.png" + }, + "class_type": "LoadImage", + "_meta": { + "title": "加载图像" + } + }, + "fields": [ + { + "field": "image", + "api_value": "1772086821573.png", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + } + ], + "source_node": { + "id": 38, + "type": "LoadImage", + "pos": [ + -86.71001308693565, + 256.1931408642567 + ], + "size": [ + 270, + 314 + ], + "flags": {}, + "order": 1, + "mode": 0, + "inputs": [ + { + "localized_name": "图像", + "name": "image", + "type": "COMBO", + "widget": { + "name": "image" + }, + "link": null + }, + { + "localized_name": "选择文件上传", + "name": "upload", + "type": "IMAGEUPLOAD", + "widget": { + "name": "upload" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 52 + ] + }, + { + "localized_name": "遮罩", + "name": "MASK", + "type": "MASK", + "links": null + } + ], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.21.1", + "Node name for S&R": "LoadImage", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [ + "1772086821573.png", + "image" + ] + }, + "status": "type_match", + "metadata": { + "source_title": "LoadImage", + "api_meta": { + "title": "加载图像" + }, + "title_equal": false + } + } + ], + "source_only_nodes": [ + { + "source_id": "31", + "node": { + "id": 31, + "type": "Reroute", + "pos": [ + 323.5439786462899, + -892.1194242532769 + ], + "size": [ + 82, + 26 + ], + "flags": {}, + "order": 3, + "mode": 0, + "inputs": [ + { + "name": "", + "type": "*", + "link": 41 + } + ], + "outputs": [ + { + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 42, + 43 + ] + } + ], + "properties": { + "showOutputText": true, + "horizontal": false, + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + } + } + }, + { + "source_id": "32", + "node": { + "id": 32, + "type": "Reroute", + "pos": [ + 305.5265691888189, + -1393.2286247891823 + ], + "size": [ + 82, + 26 + ], + "flags": {}, + "order": 4, + "mode": 0, + "inputs": [ + { + "name": "", + "type": "*", + "link": 45 + } + ], + "outputs": [ + { + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 44, + 46 + ] + } + ], + "properties": { + "showOutputText": true, + "horizontal": false, + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + } + } + }, + { + "source_id": "24", + "node": { + "id": 24, + "type": "Reroute", + "pos": [ + 238.47549649435473, + -416.21402956377005 + ], + "size": [ + 82, + 26 + ], + "flags": {}, + "order": 2, + "mode": 0, + "inputs": [ + { + "name": "", + "type": "*", + "link": 22 + } + ], + "outputs": [ + { + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 23, + 24, + 25, + 26 + ] + } + ], + "properties": { + "showOutputText": true, + "horizontal": false, + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + } + } + }, + { + "source_id": "35", + "node": { + "id": 35, + "type": "Reroute", + "pos": [ + 277.192744688726, + 23.78998347851503 + ], + "size": [ + 82, + 26 + ], + "flags": {}, + "order": 5, + "mode": 0, + "inputs": [ + { + "name": "", + "type": "*", + "link": 47 + } + ], + "outputs": [ + { + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 48 + ] + } + ], + "properties": { + "showOutputText": true, + "horizontal": false, + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + } + } + } + ], + "unconfirmed": [ + "虚拟节点、子图边界、开关与循环展开语义需独立核查", + "未有可信节点定义;单位、范围、环境和实际产物未验证" + ] + } +} diff --git a/docs/hel-892/sixth-evidence/02.json b/docs/hel-892/sixth-evidence/02.json new file mode 100644 index 00000000..7a5b0545 --- /dev/null +++ b/docs/hel-892/sixth-evidence/02.json @@ -0,0 +1,903 @@ +{ + "binding": { + "id": "02", + "source_path": "example_ComfyUI_workflows/1-语音生成/Qwen3-ASR-Subtitle语音转字幕.json", + "source_sha": "08a7519a6a6c532b8a0670dce3deff70881527ef", + "export_file": "02.api.json", + "source_sha256": "47f9afa955a6a5dce37000cb6a15bac7516d26ff50b6398bda334c825552cfd8", + "export_sha256": "daa40971b730076691721afe95e165932df94e3423f1fc44dcb7707026e7125b", + "original_filename": "Qwen3-ASR-Subtitle语音转字幕.json", + "provider": "白栖知", + "attachment_id": "01a187e2-366c-7d95-87e8-231a4b305a9e", + "received_at": "2026-10-29T13:38:38Z", + "received_at_basis": "主卡附件评论创建时间;不是导出时间", + "source_comment_id": "01a187e2-4f6e-761e-9412-43042e094daa", + "exported_at": null, + "frontend_version": null, + "node_versions": null, + "provenance": "user_supplied_api_export_environment_unknown", + "source_export_equivalence_verified": false, + "instruction": "语音转字幕:6.audio读入音频,9.audio接6的槽0;9具名hints已内联,原PrimitiveNode不在API中。1/2/3/4的source分别接9的槽0/1/2/3;原图标示STRING出口,只能确认文本预览,output_path字段不能证明已经产出字幕文件,文件映射仍待真实历史响应。" + }, + "adopted": true, + "method": "supplied_api_companion", + "general_conversion_verified": false, + "execution_verified": false, + "comparison": { + "equivalent": false, + "scope": "逐节点/字段保留API值和类型;直接连线比较ID及槽;位置控件不按顺序猜名", + "nodes": [ + { + "api_id": "1", + "api_node": { + "inputs": { + "source": [ + "9", + 3 + ] + }, + "class_type": "PreviewAny", + "_meta": { + "title": "预览任意" + } + }, + "fields": [ + { + "field": "source", + "api_value": [ + "9", + 3 + ], + "api_type": "list", + "source_link": [ + 18, + 9, + 3, + 1, + 0, + "STRING" + ], + "source_expected": [ + "9", + 3 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 1, + "type": "PreviewAny", + "pos": [ + 3410, + 5790 + ], + "size": [ + 370, + 120 + ], + "flags": {}, + "order": 6, + "mode": 0, + "inputs": [ + { + "localized_name": "源", + "name": "source", + "type": "*", + "link": 18 + } + ], + "outputs": [], + "properties": { + "Node name for S&R": "PreviewAny", + "cnr_id": "comfy-core", + "ver": "0.11.0", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.5.2" + } + }, + "widgets_values": [ + null, + null, + null + ] + }, + "status": "type_match", + "metadata": { + "source_title": "PreviewAny", + "api_meta": { + "title": "预览任意" + }, + "title_equal": false + } + }, + { + "api_id": "2", + "api_node": { + "inputs": { + "source": [ + "9", + 1 + ] + }, + "class_type": "PreviewAny", + "_meta": { + "title": "预览任意" + } + }, + "fields": [ + { + "field": "source", + "api_value": [ + "9", + 1 + ], + "api_type": "list", + "source_link": [ + 16, + 9, + 1, + 2, + 0, + "STRING" + ], + "source_expected": [ + "9", + 1 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 2, + "type": "PreviewAny", + "pos": [ + 3410, + 5410 + ], + "size": [ + 370, + 180 + ], + "flags": {}, + "order": 4, + "mode": 0, + "inputs": [ + { + "localized_name": "源", + "name": "source", + "type": "*", + "link": 16 + } + ], + "outputs": [], + "properties": { + "Node name for S&R": "PreviewAny", + "cnr_id": "comfy-core", + "ver": "0.11.0", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.5.2" + } + }, + "widgets_values": [ + null, + null, + null + ] + }, + "status": "type_match", + "metadata": { + "source_title": "PreviewAny", + "api_meta": { + "title": "预览任意" + }, + "title_equal": false + } + }, + { + "api_id": "3", + "api_node": { + "inputs": { + "source": [ + "9", + 2 + ] + }, + "class_type": "PreviewAny", + "_meta": { + "title": "预览任意" + } + }, + "fields": [ + { + "field": "source", + "api_value": [ + "9", + 2 + ], + "api_type": "list", + "source_link": [ + 17, + 9, + 2, + 3, + 0, + "STRING" + ], + "source_expected": [ + "9", + 2 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 3, + "type": "PreviewAny", + "pos": [ + 3410, + 5630 + ], + "size": [ + 370, + 120 + ], + "flags": {}, + "order": 5, + "mode": 0, + "inputs": [ + { + "localized_name": "源", + "name": "source", + "type": "*", + "link": 17 + } + ], + "outputs": [], + "properties": { + "Node name for S&R": "PreviewAny", + "cnr_id": "comfy-core", + "ver": "0.11.0", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.5.2" + } + }, + "widgets_values": [ + null, + null, + null + ] + }, + "status": "type_match", + "metadata": { + "source_title": "PreviewAny", + "api_meta": { + "title": "预览任意" + }, + "title_equal": false + } + }, + { + "api_id": "4", + "api_node": { + "inputs": { + "source": [ + "9", + 0 + ] + }, + "class_type": "PreviewAny", + "_meta": { + "title": "预览任意" + } + }, + "fields": [ + { + "field": "source", + "api_value": [ + "9", + 0 + ], + "api_type": "list", + "source_link": [ + 15, + 9, + 0, + 4, + 0, + "STRING" + ], + "source_expected": [ + "9", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 4, + "type": "PreviewAny", + "pos": [ + 3410, + 5190 + ], + "size": [ + 370, + 180 + ], + "flags": {}, + "order": 3, + "mode": 0, + "inputs": [ + { + "localized_name": "源", + "name": "source", + "type": "*", + "link": 15 + } + ], + "outputs": [], + "properties": { + "Node name for S&R": "PreviewAny", + "cnr_id": "comfy-core", + "ver": "0.11.0", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.5.2" + } + }, + "widgets_values": [ + null, + null, + null + ] + }, + "status": "type_match", + "metadata": { + "source_title": "PreviewAny", + "api_meta": { + "title": "预览任意" + }, + "title_equal": false + } + }, + { + "api_id": "6", + "api_node": { + "inputs": { + "audio": "ComfyUI-qwenasr_00003_.mp3" + }, + "class_type": "LoadAudio", + "_meta": { + "title": "加载音频" + } + }, + "fields": [ + { + "field": "audio", + "api_value": "ComfyUI-qwenasr_00003_.mp3", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + } + ], + "source_node": { + "id": 6, + "type": "LoadAudio", + "pos": [ + 2720, + 5190 + ], + "size": [ + 290, + 140 + ], + "flags": {}, + "order": 1, + "mode": 0, + "inputs": [ + { + "localized_name": "音频", + "name": "audio", + "type": "COMBO", + "widget": { + "name": "audio" + }, + "link": null + }, + { + "localized_name": "音频UI", + "name": "audioUI", + "type": "AUDIO_UI", + "widget": { + "name": "audioUI" + }, + "link": null + }, + { + "localized_name": "选择文件上传", + "name": "upload", + "type": "AUDIOUPLOAD", + "widget": { + "name": "upload" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "音频", + "name": "AUDIO", + "type": "AUDIO", + "links": [ + 13 + ] + } + ], + "properties": { + "Node name for S&R": "LoadAudio", + "cnr_id": "comfy-core", + "ver": "0.11.0", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.5.2" + } + }, + "widgets_values": [ + "ComfyUI-qwenasr_00003_.mp3", + null, + null + ] + }, + "status": "type_match", + "metadata": { + "source_title": "LoadAudio", + "api_meta": { + "title": "加载音频" + }, + "title_equal": false + } + }, + { + "api_id": "9", + "api_node": { + "inputs": { + "model": "Qwen/Qwen3-ASR-0.6B", + "precision": "bf16", + "attention": "auto", + "forced_aligner": "Qwen/Qwen3-ForcedAligner-0.6B", + "language": "auto", + "hints": "Important terms: ComfyUI, Qwen3-ASR\nUse exact casing for these terms.", + "output_format": "none", + "output_path": "", + "split_mode": "split_by_punctuation_or_pause_or_length", + "max_gap_sec": 0.6, + "max_chars": 40, + "max_inference_batch_size": 32, + "max_new_tokens": 256, + "normalize_text": true, + "unload_models": true, + "audio": [ + "6", + 0 + ] + }, + "class_type": "AILab_Qwen3ASRSubtitle", + "_meta": { + "title": "Subtitle (QwenASR)" + } + }, + "fields": [ + { + "field": "model", + "api_value": "Qwen/Qwen3-ASR-0.6B", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "precision", + "api_value": "bf16", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "attention", + "api_value": "auto", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "forced_aligner", + "api_value": "Qwen/Qwen3-ForcedAligner-0.6B", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "language", + "api_value": "auto", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "hints", + "api_value": "Important terms: ComfyUI, Qwen3-ASR\nUse exact casing for these terms.", + "api_type": "str", + "source_link": [ + 14, + 7, + 0, + 9, + 6, + "STRING" + ], + "source_expected": [ + "7", + 0 + ], + "status": "link_difference" + }, + { + "field": "output_format", + "api_value": "none", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "output_path", + "api_value": "", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "split_mode", + "api_value": "split_by_punctuation_or_pause_or_length", + "api_type": "str", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "max_gap_sec", + "api_value": 0.6, + "api_type": "float", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "max_chars", + "api_value": 40, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "max_inference_batch_size", + "api_value": 32, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "max_new_tokens", + "api_value": 256, + "api_type": "int", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "normalize_text", + "api_value": true, + "api_type": "bool", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "unload_models", + "api_value": true, + "api_type": "bool", + "status": "unnamed_or_virtual_boundary_unverified" + }, + { + "field": "audio", + "api_value": [ + "6", + 0 + ], + "api_type": "list", + "source_link": [ + 13, + 6, + 0, + 9, + 0, + "AUDIO" + ], + "source_expected": [ + "6", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 9, + "type": "AILab_Qwen3ASRSubtitle", + "pos": [ + 3030, + 5190 + ], + "size": [ + 340, + 460 + ], + "flags": {}, + "order": 2, + "mode": 0, + "inputs": [ + { + "localized_name": "audio", + "name": "audio", + "type": "AUDIO", + "link": 13 + }, + { + "localized_name": "model", + "name": "model", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "model" + }, + "link": null + }, + { + "localized_name": "precision", + "name": "precision", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "precision" + }, + "link": null + }, + { + "localized_name": "attention", + "name": "attention", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "attention" + }, + "link": null + }, + { + "localized_name": "forced_aligner", + "name": "forced_aligner", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "forced_aligner" + }, + "link": null + }, + { + "localized_name": "language", + "name": "language", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "language" + }, + "link": null + }, + { + "localized_name": "hints", + "name": "hints", + "shape": 7, + "type": "STRING", + "widget": { + "name": "hints" + }, + "link": 14 + }, + { + "localized_name": "output_format", + "name": "output_format", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "output_format" + }, + "link": null + }, + { + "localized_name": "output_path", + "name": "output_path", + "shape": 7, + "type": "STRING", + "widget": { + "name": "output_path" + }, + "link": null + }, + { + "localized_name": "split_mode", + "name": "split_mode", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "split_mode" + }, + "link": null + }, + { + "localized_name": "max_gap_sec", + "name": "max_gap_sec", + "shape": 7, + "type": "FLOAT", + "widget": { + "name": "max_gap_sec" + }, + "link": null + }, + { + "localized_name": "max_chars", + "name": "max_chars", + "shape": 7, + "type": "INT", + "widget": { + "name": "max_chars" + }, + "link": null + }, + { + "localized_name": "max_inference_batch_size", + "name": "max_inference_batch_size", + "shape": 7, + "type": "INT", + "widget": { + "name": "max_inference_batch_size" + }, + "link": null + }, + { + "localized_name": "max_new_tokens", + "name": "max_new_tokens", + "shape": 7, + "type": "INT", + "widget": { + "name": "max_new_tokens" + }, + "link": null + }, + { + "localized_name": "unload_models", + "name": "unload_models", + "shape": 7, + "type": "BOOLEAN", + "widget": { + "name": "unload_models" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "TEXT", + "name": "TEXT", + "type": "STRING", + "links": [ + 15 + ] + }, + { + "localized_name": "SUBTITLES", + "name": "SUBTITLES", + "type": "STRING", + "links": [ + 16 + ] + }, + { + "localized_name": "LANUGAGE", + "name": "LANUGAGE", + "type": "STRING", + "links": [ + 17 + ] + }, + { + "localized_name": "OUTPUT_PATH", + "name": "OUTPUT_PATH", + "type": "STRING", + "links": [ + 18 + ] + } + ], + "properties": { + "Node name for S&R": "AILab_Qwen3ASRSubtitle", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [ + "Qwen/Qwen3-ASR-0.6B", + "bf16", + "auto", + "Qwen/Qwen3-ForcedAligner-0.6B", + "auto", + "Important terms: ComfyUI, Qwen3-ASR\nUse exact casing for these terms.", + "none", + "", + "split_by_punctuation_or_pause_or_length", + 0.6, + 40, + 32, + 256, + true + ], + "color": "#151c19", + "bgcolor": "#444d49" + }, + "status": "type_match", + "metadata": { + "source_title": "AILab_Qwen3ASRSubtitle", + "api_meta": { + "title": "Subtitle (QwenASR)" + }, + "title_equal": false + } + } + ], + "source_only_nodes": [ + { + "source_id": "7", + "node": { + "id": 7, + "type": "PrimitiveNode", + "pos": [ + 2720, + 5380 + ], + "size": [ + 290, + 90 + ], + "flags": {}, + "order": 0, + "mode": 0, + "inputs": [], + "outputs": [ + { + "name": "STRING", + "type": "STRING", + "widget": { + "name": "hints" + }, + "links": [ + 14 + ] + } + ], + "properties": { + "Run widget replace on values": false, + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "Important terms: ComfyUI, Qwen3-ASR\nUse exact casing for these terms." + ] + } + } + ], + "unconfirmed": [ + "虚拟节点、子图边界、开关与循环展开语义需独立核查", + "未有可信节点定义;单位、范围、环境和实际产物未验证" + ] + } +} diff --git a/docs/hel-892/sixth-evidence/03.json b/docs/hel-892/sixth-evidence/03.json new file mode 100644 index 00000000..4acba768 --- /dev/null +++ b/docs/hel-892/sixth-evidence/03.json @@ -0,0 +1,2211 @@ +{ + "binding": { + "id": "03", + "source_path": "example_ComfyUI_workflows/2-音乐生成/minimax_music_文生音乐.json", + "source_sha": "c248b9f72ba42a8fe2e27ab3762d2ef56869b0f8", + "export_file": "03.api.json", + "source_sha256": "0a655ead5a0da406159dceb46451439df0c02050becaaa25b3018abb41840284", + "export_sha256": "91e6c2cea3747a03bbb71f661dbd016b28040781482b956365ff76a7a8b647ce", + "original_filename": "minimax_music_文生音乐.json", + "provider": "白栖知", + "attachment_id": "01a187e2-366a-7575-b7e1-f4d66e5e2c1d", + "received_at": "2026-10-29T13:38:38Z", + "received_at_basis": "主卡附件评论创建时间;不是导出时间", + "source_comment_id": "01a187e2-4f6e-761e-9412-43042e094daa", + "exported_at": null, + "frontend_version": null, + "node_versions": null, + "provenance": "user_supplied_api_export_environment_unknown", + "source_export_equivalence_verified": false, + "instruction": "音乐:37子图展开为37:*节点;37:13.caption/lyrics经clip编码,37:38.seed连接编码器和采样器;37:43.switch选择普通/分块解码,35.audio接开关结果并配置filename_prefix/format/format.quality。先核对保存节点、时长及开关现值;尚未验证音频文件响应。" + }, + "adopted": true, + "method": "supplied_api_companion", + "general_conversion_verified": false, + "execution_verified": false, + "comparison": { + "equivalent": false, + "scope": "逐节点/字段保留API值和类型;直接连线比较ID及槽;位置控件不按顺序猜名", + "nodes": [ + { + "api_id": "35", + "api_node": { + "inputs": { + "filename_prefix": "audio/audio_minimax_music3", + "format": "mp3", + "format.quality": "V0", + "audio": [ + "37:43", + 0 + ] + }, + "class_type": "SaveAudioAdvanced", + "_meta": { + "title": "保存音频(高级)" + } + }, + "fields": [ + { + "field": "filename_prefix", + "api_value": "audio/audio_minimax_music3", + "api_type": "str", + "named_evidence": { + "source_value": "audio/audio_minimax_music3", + "source_type": "str", + "status": "equal" + }, + "source_value": "audio/audio_minimax_music3", + "status": "named_equal" + }, + { + "field": "format", + "api_value": "mp3", + "api_type": "str", + "named_evidence": { + "source_value": "mp3", + "source_type": "str", + "status": "equal" + }, + "source_value": "mp3", + "status": "named_equal" + }, + { + "field": "format.quality", + "api_value": "V0", + "api_type": "str", + "named_evidence": { + "source_value": "V0", + "source_type": "str", + "status": "equal" + }, + "source_value": "V0", + "status": "named_equal" + }, + { + "field": "audio", + "api_value": [ + "37:43", + 0 + ], + "api_type": "list", + "source_link": [ + 42, + 37, + 0, + 35, + 0, + "AUDIO" + ], + "source_expected": [ + "37", + 0 + ], + "status": "link_difference" + } + ], + "source_node": { + "id": 35, + "type": "SaveAudioAdvanced", + "pos": [ + 2150, + 630 + ], + "size": [ + 750, + 160 + ], + "flags": {}, + "order": 3, + "mode": 0, + "inputs": [ + { + "localized_name": "audio", + "name": "audio", + "type": "AUDIO", + "link": 42 + }, + { + "localized_name": "filename_prefix", + "name": "filename_prefix", + "type": "STRING", + "widget": { + "name": "filename_prefix" + }, + "link": null + }, + { + "localized_name": "format", + "name": "format", + "type": "COMFY_DYNAMICCOMBO_V3", + "widget": { + "name": "format" + }, + "link": null + }, + { + "localized_name": "format.quality", + "name": "format.quality", + "type": "COMBO", + "widget": { + "name": "format.quality" + }, + "link": null + }, + { + "localized_name": "audioUI", + "name": "audioUI", + "type": "AUDIO_UI", + "widget": { + "name": "audioUI" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "audio", + "name": "audio", + "type": "AUDIO", + "links": null + } + ], + "properties": { + "Node name for S&R": "SaveAudioAdvanced", + "cnr_id": "comfy-core", + "ver": "0.31.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "audio/audio_minimax_music3", + "mp3", + "V0" + ], + "widgets_values_named": { + "filename_prefix": "audio/audio_minimax_music3", + "format": "mp3", + "format.quality": "V0" + } + }, + "status": "type_match", + "metadata": { + "source_title": "SaveAudioAdvanced", + "api_meta": { + "title": "保存音频(高级)" + }, + "title_equal": false + } + }, + { + "api_id": "37:6", + "api_node": { + "inputs": { + "unet_name": "MiniMax/minimax_music3_dit_fp16.safetensors", + "weight_dtype": "default" + }, + "class_type": "UNETLoader", + "_meta": { + "title": "UNet加载器" + } + }, + "fields": [ + { + "field": "unet_name", + "api_value": "MiniMax/minimax_music3_dit_fp16.safetensors", + "api_type": "str", + "named_evidence": { + "source_value": "minimax_music3_dit_fp16.safetensors", + "source_type": "str", + "status": "different" + }, + "source_link": [ + 55, + -10, + 4, + 6, + 0, + "COMBO" + ], + "source_expected": [ + "37:-10", + 4 + ], + "status": "link_difference" + }, + { + "field": "weight_dtype", + "api_value": "default", + "api_type": "str", + "named_evidence": { + "source_value": "default", + "source_type": "str", + "status": "equal" + }, + "source_value": "default", + "status": "named_equal" + } + ], + "source_node": { + "id": 6, + "type": "UNETLoader", + "pos": [ + 700, + 380 + ], + "size": [ + 480, + 90 + ], + "flags": {}, + "order": 1, + "mode": 0, + "showAdvanced": false, + "inputs": [ + { + "localized_name": "UNet名称", + "name": "unet_name", + "type": "COMBO", + "widget": { + "name": "unet_name" + }, + "link": 55 + }, + { + "localized_name": "数据类型", + "name": "weight_dtype", + "type": "COMBO", + "widget": { + "name": "weight_dtype" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "模型", + "name": "MODEL", + "type": "MODEL", + "links": [ + 6 + ] + } + ], + "properties": { + "Node name for S&R": "UNETLoader", + "cnr_id": "comfy-core", + "ver": "0.31.0", + "models": [ + { + "name": "minimax_music3_dit_fp16.safetensors", + "url": "https://huggingface.co/Comfy-Org/MiniMax-Music-3/resolve/main/diffusion_models/minimax_music3_dit_fp16.safetensors", + "directory": "diffusion_models" + } + ], + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "minimax_music3_dit_fp16.safetensors", + "default" + ], + "widgets_values_named": { + "unet_name": "minimax_music3_dit_fp16.safetensors", + "weight_dtype": "default" + } + }, + "status": "type_match", + "metadata": { + "source_title": "UNETLoader", + "api_meta": { + "title": "UNet加载器" + }, + "title_equal": false + } + }, + { + "api_id": "37:13", + "api_node": { + "inputs": { + "caption": "Global Metadata: Lo-fi hip-hop, chillhop. 78 BPM, D flat major, major scale with jazzy extensions. Laid-back and dreamy throughout, a gentle warm drift with a subtle late-night glow that deepens in the middle and dissolves softly at the end. Studying, raining-outside, headphones-on late-night listening. Bedroom production: muddy warm texture, heavy vinyl crackle, tape hiss and wow-flutter pitch wobble, low-passed dusty mix, soft-clipped drums, everything slightly detuned and cozy.\n\nVocal Details: Soft androgynous vocal, hushed half-sung half-spoken delivery, sitting low in the mix like another instrument, lazy behind-the-beat phrasing, gentle breathy timbre. Sparse murmured double-tracked harmonies, occasional wordless \"mmm\" and \"ooh\" hums drenched in tape delay and warm spring reverb. Long stretches with no vocals at all.\n\nArrangement: Dusty boom-bap drums with a soft thumping kick, cracked snare with lazy swing, brushed hi-hats, low round sub bass. Warm Rhodes piano chords with slow chorus wobble as the harmonic bed, mellow jazzy guitar licks answering the vocal lines, constant vinyl crackle as texture. Intro: rain and vinyl noise, solo Rhodes chords fading in, drums slipping in halfway. Verses: minimal — drums, bass, Rhodes, soft guitar fills between lines. Instrumental sections: guitar and Rhodes trade relaxed jazzy phrases over the beat, occasional muted trumpet ghost notes far in the background. Bridge: drums drop away to rain, crackle, and floating detuned Rhodes, then the beat eases back in. Outro: elements fade one by one until only vinyl crackle and a last unresolved Rhodes chord remain.", + "lyrics": "[Intro]\nMmm...\n(rain on the window)\nOoh...\n\n[Verse]\nMidnight and the canvas glows\nDragging little wires where the current flows\nType a quiet dream, let the sampler drift\nNoise into a picture, like the fog just lifts\nTwenty slow steps, I'm in no hurry now\nLatents turning colors and I don't know how\nEvery render's like a polaroid I found\nSoft focus memories, no sound\n\n[Instrumental]\n\n[Verse]\nQueue another frame, let the motion breathe\nPictures start to move like the falling leaves\nVideo drifting by at twenty-four\nLittle animations on my bedroom floor\nSeed after seed like the rain outside\nSome of them are keepers, some I let slide\nSave the ones that feel like a Sunday slow\nNode to node to node... and off we go\n\n[Chorus]\nMmm... let it render on\n(take your time, take your time)\nOoh... by the morning it'll all be done\n(one more queue, one more try)\n\n[Instrumental]\n\n[Bridge]\nRain keeps drawing pictures on the glass...\nMy machine keeps dreaming...\nNeither of us fast...\n\n[Chorus]\nMmm... let it render on\n(take your time, take your time)\nOoh... by the morning it'll all be done\n(one more queue, one more try)\n\n[Instrumental]\n\n[Outro]\nMmm...\n(node to node)\nOoh... goodnight", + "seed": [ + "37:38", + 0 + ], + "max_duration": 60, + "cfg_scale": 1.7, + "top_k": 50, + "clip": [ + "37:3", + 0 + ] + }, + "class_type": "MiniMaxMusic3TextEncode", + "_meta": { + "title": "MiniMax Music3 文本编码" + } + }, + "fields": [ + { + "field": "caption", + "api_value": "Global Metadata: Lo-fi hip-hop, chillhop. 78 BPM, D flat major, major scale with jazzy extensions. Laid-back and dreamy throughout, a gentle warm drift with a subtle late-night glow that deepens in the middle and dissolves softly at the end. Studying, raining-outside, headphones-on late-night listening. Bedroom production: muddy warm texture, heavy vinyl crackle, tape hiss and wow-flutter pitch wobble, low-passed dusty mix, soft-clipped drums, everything slightly detuned and cozy.\n\nVocal Details: Soft androgynous vocal, hushed half-sung half-spoken delivery, sitting low in the mix like another instrument, lazy behind-the-beat phrasing, gentle breathy timbre. Sparse murmured double-tracked harmonies, occasional wordless \"mmm\" and \"ooh\" hums drenched in tape delay and warm spring reverb. Long stretches with no vocals at all.\n\nArrangement: Dusty boom-bap drums with a soft thumping kick, cracked snare with lazy swing, brushed hi-hats, low round sub bass. Warm Rhodes piano chords with slow chorus wobble as the harmonic bed, mellow jazzy guitar licks answering the vocal lines, constant vinyl crackle as texture. Intro: rain and vinyl noise, solo Rhodes chords fading in, drums slipping in halfway. Verses: minimal — drums, bass, Rhodes, soft guitar fills between lines. Instrumental sections: guitar and Rhodes trade relaxed jazzy phrases over the beat, occasional muted trumpet ghost notes far in the background. Bridge: drums drop away to rain, crackle, and floating detuned Rhodes, then the beat eases back in. Outro: elements fade one by one until only vinyl crackle and a last unresolved Rhodes chord remain.", + "api_type": "str", + "named_evidence": { + "source_value": "", + "source_type": "str", + "status": "different" + }, + "source_link": [ + 45, + -10, + 0, + 13, + 1, + "STRING" + ], + "source_expected": [ + "37:-10", + 0 + ], + "status": "link_difference" + }, + { + "field": "lyrics", + "api_value": "[Intro]\nMmm...\n(rain on the window)\nOoh...\n\n[Verse]\nMidnight and the canvas glows\nDragging little wires where the current flows\nType a quiet dream, let the sampler drift\nNoise into a picture, like the fog just lifts\nTwenty slow steps, I'm in no hurry now\nLatents turning colors and I don't know how\nEvery render's like a polaroid I found\nSoft focus memories, no sound\n\n[Instrumental]\n\n[Verse]\nQueue another frame, let the motion breathe\nPictures start to move like the falling leaves\nVideo drifting by at twenty-four\nLittle animations on my bedroom floor\nSeed after seed like the rain outside\nSome of them are keepers, some I let slide\nSave the ones that feel like a Sunday slow\nNode to node to node... and off we go\n\n[Chorus]\nMmm... let it render on\n(take your time, take your time)\nOoh... by the morning it'll all be done\n(one more queue, one more try)\n\n[Instrumental]\n\n[Bridge]\nRain keeps drawing pictures on the glass...\nMy machine keeps dreaming...\nNeither of us fast...\n\n[Chorus]\nMmm... let it render on\n(take your time, take your time)\nOoh... by the morning it'll all be done\n(one more queue, one more try)\n\n[Instrumental]\n\n[Outro]\nMmm...\n(node to node)\nOoh... goodnight", + "api_type": "str", + "named_evidence": { + "source_value": "", + "source_type": "str", + "status": "different" + }, + "source_link": [ + 46, + -10, + 1, + 13, + 2, + "STRING" + ], + "source_expected": [ + "37:-10", + 1 + ], + "status": "link_difference" + }, + { + "field": "seed", + "api_value": [ + "37:38", + 0 + ], + "api_type": "list", + "named_evidence": { + "source_value": 222, + "source_type": "int", + "status": "different" + }, + "source_link": [ + 52, + 38, + 0, + 13, + 3, + "INT" + ], + "source_expected": [ + "37:38", + 0 + ], + "status": "link_equal" + }, + { + "field": "max_duration", + "api_value": 60, + "api_type": "int", + "named_evidence": { + "source_value": 60, + "source_type": "int", + "status": "equal" + }, + "source_link": [ + 47, + -10, + 2, + 13, + 4, + "FLOAT" + ], + "source_expected": [ + "37:-10", + 2 + ], + "status": "link_difference" + }, + { + "field": "cfg_scale", + "api_value": 1.7, + "api_type": "float", + "named_evidence": { + "source_value": 1.7, + "source_type": "float", + "status": "equal" + }, + "source_value": 1.7, + "status": "named_equal" + }, + { + "field": "top_k", + "api_value": 50, + "api_type": "int", + "named_evidence": { + "source_value": 50, + "source_type": "int", + "status": "equal" + }, + "source_value": 50, + "status": "named_equal" + }, + { + "field": "clip", + "api_value": [ + "37:3", + 0 + ], + "api_type": "list", + "source_link": [ + 18, + 3, + 0, + 13, + 0, + "CLIP" + ], + "source_expected": [ + "37:3", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 13, + "type": "MiniMaxMusic3TextEncode", + "pos": [ + 1230, + 380 + ], + "size": [ + 500, + 750 + ], + "flags": {}, + "order": 6, + "mode": 0, + "showAdvanced": true, + "inputs": [ + { + "localized_name": "clip", + "name": "clip", + "type": "CLIP", + "link": 18 + }, + { + "localized_name": "caption", + "name": "caption", + "type": "STRING", + "widget": { + "name": "caption" + }, + "link": 45 + }, + { + "localized_name": "lyrics", + "name": "lyrics", + "type": "STRING", + "widget": { + "name": "lyrics" + }, + "link": 46 + }, + { + "localized_name": "seed", + "name": "seed", + "type": "INT", + "widget": { + "name": "seed" + }, + "link": 52 + }, + { + "localized_name": "max_duration", + "name": "max_duration", + "type": "FLOAT", + "widget": { + "name": "max_duration" + }, + "link": 47 + }, + { + "localized_name": "cfg_scale", + "name": "cfg_scale", + "type": "FLOAT", + "widget": { + "name": "cfg_scale" + }, + "link": null + }, + { + "localized_name": "top_k", + "name": "top_k", + "type": "INT", + "widget": { + "name": "top_k" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "条件", + "name": "CONDITIONING", + "type": "CONDITIONING", + "links": [ + 15, + 16 + ] + }, + { + "localized_name": "seconds", + "name": "seconds", + "type": "FLOAT", + "links": [ + 20 + ] + } + ], + "properties": { + "Node name for S&R": "MiniMaxMusic3TextEncode", + "cnr_id": "comfy-core", + "ver": "0.31.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "", + "", + 222, + "fixed", + 60, + 1.7, + 50 + ], + "widgets_values_named": { + "caption": "", + "lyrics": "", + "seed": 222, + "control_after_generate": "fixed", + "max_duration": 60, + "cfg_scale": 1.7, + "top_k": 50 + } + }, + "status": "type_match", + "metadata": { + "source_title": "MiniMaxMusic3TextEncode", + "api_meta": { + "title": "MiniMax Music3 文本编码" + }, + "title_equal": false + } + }, + { + "api_id": "37:3", + "api_node": { + "inputs": { + "clip_name": "MiniMax/minimax_music3_text_encoder_pruned_int8_convrot.safetensors", + "type": "minimax", + "device": "default" + }, + "class_type": "CLIPLoader", + "_meta": { + "title": "加载CLIP" + } + }, + "fields": [ + { + "field": "clip_name", + "api_value": "MiniMax/minimax_music3_text_encoder_pruned_int8_convrot.safetensors", + "api_type": "str", + "named_evidence": { + "source_value": "minimax_music3_text_encoder_pruned_int8_convrot.safetensors", + "source_type": "str", + "status": "different" + }, + "source_link": [ + 56, + -10, + 5, + 3, + 0, + "COMBO" + ], + "source_expected": [ + "37:-10", + 5 + ], + "status": "link_difference" + }, + { + "field": "type", + "api_value": "minimax", + "api_type": "str", + "named_evidence": { + "source_value": "minimax", + "source_type": "str", + "status": "equal" + }, + "source_value": "minimax", + "status": "named_equal" + }, + { + "field": "device", + "api_value": "default", + "api_type": "str", + "named_evidence": { + "source_value": "default", + "source_type": "str", + "status": "equal" + }, + "source_value": "default", + "status": "named_equal" + } + ], + "source_node": { + "id": 3, + "type": "CLIPLoader", + "pos": [ + 700, + 550 + ], + "size": [ + 480, + 120 + ], + "flags": {}, + "order": 0, + "mode": 0, + "inputs": [ + { + "localized_name": "CLIP名称", + "name": "clip_name", + "type": "COMBO", + "widget": { + "name": "clip_name" + }, + "link": 56 + }, + { + "localized_name": "类型", + "name": "type", + "type": "COMBO", + "widget": { + "name": "type" + }, + "link": null + }, + { + "localized_name": "设备", + "name": "device", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "device" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "CLIP", + "name": "CLIP", + "type": "CLIP", + "links": [ + 18 + ] + } + ], + "properties": { + "Node name for S&R": "CLIPLoader", + "cnr_id": "comfy-core", + "ver": "0.31.0", + "models": [ + { + "name": "minimax_music3_text_encoder_pruned_int8_convrot.safetensors", + "url": "https://huggingface.co/Comfy-Org/MiniMax-Music-3/resolve/main/text_encoders/minimax_music3_text_encoder_pruned_int8_convrot.safetensors", + "directory": "text_encoders" + } + ], + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "minimax_music3_text_encoder_pruned_int8_convrot.safetensors", + "minimax", + "default" + ], + "widgets_values_named": { + "clip_name": "minimax_music3_text_encoder_pruned_int8_convrot.safetensors", + "type": "minimax", + "device": "default" + } + }, + "status": "type_match", + "metadata": { + "source_title": "CLIPLoader", + "api_meta": { + "title": "加载CLIP" + }, + "title_equal": false + } + }, + { + "api_id": "37:7", + "api_node": { + "inputs": { + "vae_name": "MiniMax/minimax_music3_dav.safetensors" + }, + "class_type": "VAELoader", + "_meta": { + "title": "加载VAE" + } + }, + "fields": [ + { + "field": "vae_name", + "api_value": "MiniMax/minimax_music3_dav.safetensors", + "api_type": "str", + "named_evidence": { + "source_value": "minimax_music3_dav.safetensors", + "source_type": "str", + "status": "different" + }, + "source_link": [ + 57, + -10, + 6, + 7, + 0, + "COMBO" + ], + "source_expected": [ + "37:-10", + 6 + ], + "status": "link_difference" + } + ], + "source_node": { + "id": 7, + "type": "VAELoader", + "pos": [ + 700, + 750 + ], + "size": [ + 480, + 70 + ], + "flags": {}, + "order": 2, + "mode": 0, + "inputs": [ + { + "localized_name": "vae名称", + "name": "vae_name", + "type": "COMBO", + "widget": { + "name": "vae_name" + }, + "link": 57 + } + ], + "outputs": [ + { + "localized_name": "VAE", + "name": "VAE", + "type": "VAE", + "links": [ + 13, + 59 + ] + } + ], + "properties": { + "Node name for S&R": "VAELoader", + "cnr_id": "comfy-core", + "ver": "0.31.0", + "models": [ + { + "name": "minimax_music3_dav.safetensors", + "url": "https://huggingface.co/Comfy-Org/MiniMax-Music-3/resolve/main/vae/minimax_music3_dav.safetensors", + "directory": "vae" + } + ], + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "minimax_music3_dav.safetensors" + ], + "widgets_values_named": { + "vae_name": "minimax_music3_dav.safetensors" + } + }, + "status": "type_match", + "metadata": { + "source_title": "VAELoader", + "api_meta": { + "title": "加载VAE" + }, + "title_equal": false + } + }, + { + "api_id": "37:10", + "api_node": { + "inputs": { + "conditioning": [ + "37:13", + 0 + ] + }, + "class_type": "ConditioningZeroOut", + "_meta": { + "title": "条件零化" + } + }, + "fields": [ + { + "field": "conditioning", + "api_value": [ + "37:13", + 0 + ], + "api_type": "list", + "source_link": [ + 16, + 13, + 0, + 10, + 0, + "CONDITIONING" + ], + "source_expected": [ + "37:13", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 10, + "type": "ConditioningZeroOut", + "pos": [ + 1820, + 540 + ], + "size": [ + 230, + 40 + ], + "flags": {}, + "order": 4, + "mode": 0, + "inputs": [ + { + "localized_name": "条件", + "name": "conditioning", + "type": "CONDITIONING", + "link": 16 + } + ], + "outputs": [ + { + "localized_name": "条件", + "name": "CONDITIONING", + "type": "CONDITIONING", + "links": [ + 8 + ] + } + ], + "properties": { + "Node name for S&R": "ConditioningZeroOut", + "cnr_id": "comfy-core", + "ver": "0.31.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + } + }, + "status": "type_match", + "metadata": { + "source_title": "ConditioningZeroOut", + "api_meta": { + "title": "条件零化" + }, + "title_equal": false + } + }, + { + "api_id": "37:15", + "api_node": { + "inputs": { + "seconds": [ + "37:13", + 1 + ], + "batch_size": 1 + }, + "class_type": "EmptyMiniMaxMusic3LatentAudio", + "_meta": { + "title": "空 MiniMax Music3 潜空间音频" + } + }, + "fields": [ + { + "field": "seconds", + "api_value": [ + "37:13", + 1 + ], + "api_type": "list", + "named_evidence": { + "source_value": 120, + "source_type": "int", + "status": "different" + }, + "source_link": [ + 20, + 13, + 1, + 15, + 0, + "FLOAT" + ], + "source_expected": [ + "37:13", + 1 + ], + "status": "link_equal" + }, + { + "field": "batch_size", + "api_value": 1, + "api_type": "int", + "named_evidence": { + "source_value": 1, + "source_type": "int", + "status": "equal" + }, + "source_value": 1, + "status": "named_equal" + } + ], + "source_node": { + "id": 15, + "type": "EmptyMiniMaxMusic3LatentAudio", + "pos": [ + 1780, + 380 + ], + "size": [ + 350, + 90 + ], + "flags": {}, + "order": 7, + "mode": 0, + "inputs": [ + { + "localized_name": "seconds", + "name": "seconds", + "type": "FLOAT", + "widget": { + "name": "seconds" + }, + "link": 20 + }, + { + "localized_name": "batch_size", + "name": "batch_size", + "type": "INT", + "widget": { + "name": "batch_size" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "Latent", + "name": "LATENT", + "type": "LATENT", + "links": [ + 21 + ] + } + ], + "properties": { + "Node name for S&R": "EmptyMiniMaxMusic3LatentAudio", + "cnr_id": "comfy-core", + "ver": "0.31.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + 120, + 1 + ], + "widgets_values_named": { + "seconds": 120, + "batch_size": 1 + } + }, + "status": "type_match", + "metadata": { + "source_title": "EmptyMiniMaxMusic3LatentAudio", + "api_meta": { + "title": "空 MiniMax Music3 潜空间音频" + }, + "title_equal": false + } + }, + { + "api_id": "37:9", + "api_node": { + "inputs": { + "seed": [ + "37:38", + 0 + ], + "steps": 30, + "cfg": 1.7, + "sampler_name": "euler", + "scheduler": "simple", + "denoise": 1, + "model": [ + "37:6", + 0 + ], + "positive": [ + "37:13", + 0 + ], + "negative": [ + "37:10", + 0 + ], + "latent_image": [ + "37:15", + 0 + ] + }, + "class_type": "KSampler", + "_meta": { + "title": "K采样器" + } + }, + "fields": [ + { + "field": "seed", + "api_value": [ + "37:38", + 0 + ], + "api_type": "list", + "named_evidence": { + "source_value": 1111111112, + "source_type": "int", + "status": "different" + }, + "source_link": [ + 53, + 38, + 0, + 9, + 4, + "INT" + ], + "source_expected": [ + "37:38", + 0 + ], + "status": "link_equal" + }, + { + "field": "steps", + "api_value": 30, + "api_type": "int", + "named_evidence": { + "source_value": 30, + "source_type": "int", + "status": "equal" + }, + "source_value": 30, + "status": "named_equal" + }, + { + "field": "cfg", + "api_value": 1.7, + "api_type": "float", + "named_evidence": { + "source_value": 1.7, + "source_type": "float", + "status": "equal" + }, + "source_value": 1.7, + "status": "named_equal" + }, + { + "field": "sampler_name", + "api_value": "euler", + "api_type": "str", + "named_evidence": { + "source_value": "euler", + "source_type": "str", + "status": "equal" + }, + "source_value": "euler", + "status": "named_equal" + }, + { + "field": "scheduler", + "api_value": "simple", + "api_type": "str", + "named_evidence": { + "source_value": "simple", + "source_type": "str", + "status": "equal" + }, + "source_value": "simple", + "status": "named_equal" + }, + { + "field": "denoise", + "api_value": 1, + "api_type": "int", + "named_evidence": { + "source_value": 1, + "source_type": "int", + "status": "equal" + }, + "source_value": 1, + "status": "named_equal" + }, + { + "field": "model", + "api_value": [ + "37:6", + 0 + ], + "api_type": "list", + "source_link": [ + 6, + 6, + 0, + 9, + 0, + "MODEL" + ], + "source_expected": [ + "37:6", + 0 + ], + "status": "link_equal" + }, + { + "field": "positive", + "api_value": [ + "37:13", + 0 + ], + "api_type": "list", + "source_link": [ + 15, + 13, + 0, + 9, + 1, + "CONDITIONING" + ], + "source_expected": [ + "37:13", + 0 + ], + "status": "link_equal" + }, + { + "field": "negative", + "api_value": [ + "37:10", + 0 + ], + "api_type": "list", + "source_link": [ + 8, + 10, + 0, + 9, + 2, + "CONDITIONING" + ], + "source_expected": [ + "37:10", + 0 + ], + "status": "link_equal" + }, + { + "field": "latent_image", + "api_value": [ + "37:15", + 0 + ], + "api_type": "list", + "source_link": [ + 21, + 15, + 0, + 9, + 3, + "LATENT" + ], + "source_expected": [ + "37:15", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 9, + "type": "KSampler", + "pos": [ + 2170, + 380 + ], + "size": [ + 370, + 270 + ], + "flags": {}, + "order": 3, + "mode": 0, + "inputs": [ + { + "localized_name": "模型", + "name": "model", + "type": "MODEL", + "link": 6 + }, + { + "localized_name": "正面条件", + "name": "positive", + "type": "CONDITIONING", + "link": 15 + }, + { + "localized_name": "负面条件", + "name": "negative", + "type": "CONDITIONING", + "link": 8 + }, + { + "localized_name": "Latent图像", + "name": "latent_image", + "type": "LATENT", + "link": 21 + }, + { + "localized_name": "种子", + "name": "seed", + "type": "INT", + "widget": { + "name": "seed" + }, + "link": 53 + }, + { + "localized_name": "步数", + "name": "steps", + "type": "INT", + "widget": { + "name": "steps" + }, + "link": null + }, + { + "localized_name": "cfg", + "name": "cfg", + "type": "FLOAT", + "widget": { + "name": "cfg" + }, + "link": null + }, + { + "localized_name": "采样器名称", + "name": "sampler_name", + "type": "COMBO", + "widget": { + "name": "sampler_name" + }, + "link": null + }, + { + "localized_name": "调度器", + "name": "scheduler", + "type": "COMBO", + "widget": { + "name": "scheduler" + }, + "link": null + }, + { + "localized_name": "降噪", + "name": "denoise", + "type": "FLOAT", + "widget": { + "name": "denoise" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "Latent", + "name": "LATENT", + "type": "LATENT", + "links": [ + 12, + 58 + ] + } + ], + "properties": { + "Node name for S&R": "KSampler", + "cnr_id": "comfy-core", + "ver": "0.31.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + 1111111112, + "fixed", + 30, + 1.7, + "euler", + "simple", + 1 + ], + "widgets_values_named": { + "seed": 1111111112, + "control_after_generate": "fixed", + "steps": 30, + "cfg": 1.7, + "sampler_name": "euler", + "scheduler": "simple", + "denoise": 1 + } + }, + "status": "type_match", + "metadata": { + "source_title": "KSampler", + "api_meta": { + "title": "K采样器" + }, + "title_equal": false + } + }, + { + "api_id": "37:12", + "api_node": { + "inputs": { + "samples": [ + "37:9", + 0 + ], + "vae": [ + "37:7", + 0 + ] + }, + "class_type": "VAEDecodeAudio", + "_meta": { + "title": "VAE解码(音频)" + } + }, + "fields": [ + { + "field": "samples", + "api_value": [ + "37:9", + 0 + ], + "api_type": "list", + "source_link": [ + 12, + 9, + 0, + 12, + 0, + "LATENT" + ], + "source_expected": [ + "37:9", + 0 + ], + "status": "link_equal" + }, + { + "field": "vae", + "api_value": [ + "37:7", + 0 + ], + "api_type": "list", + "source_link": [ + 13, + 7, + 0, + 12, + 1, + "VAE" + ], + "source_expected": [ + "37:7", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 12, + "type": "VAEDecodeAudio", + "pos": [ + 2590, + 380 + ], + "size": [ + 230, + 60 + ], + "flags": {}, + "order": 5, + "mode": 0, + "inputs": [ + { + "localized_name": "Latent", + "name": "samples", + "type": "LATENT", + "link": 12 + }, + { + "localized_name": "vae", + "name": "vae", + "type": "VAE", + "link": 13 + } + ], + "outputs": [ + { + "localized_name": "音频", + "name": "AUDIO", + "type": "AUDIO", + "links": [ + 60 + ] + } + ], + "properties": { + "Node name for S&R": "VAEDecodeAudio", + "cnr_id": "comfy-core", + "ver": "0.31.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + } + }, + "status": "type_match", + "metadata": { + "source_title": "VAEDecodeAudio", + "api_meta": { + "title": "VAE解码(音频)" + }, + "title_equal": false + } + }, + { + "api_id": "37:38", + "api_node": { + "inputs": { + "seed": 617970236276137 + }, + "class_type": "SeedNode", + "_meta": { + "title": "种子" + } + }, + "fields": [ + { + "field": "seed", + "api_value": 617970236276137, + "api_type": "int", + "named_evidence": { + "source_value": 0, + "source_type": "int", + "status": "different" + }, + "source_link": [ + 54, + -10, + 3, + 38, + 0, + "INT" + ], + "source_expected": [ + "37:-10", + 3 + ], + "status": "link_difference" + } + ], + "source_node": { + "id": 38, + "type": "SeedNode", + "pos": [ + 890, + 970 + ], + "size": [ + 270, + 90 + ], + "flags": {}, + "order": 8, + "mode": 0, + "inputs": [ + { + "localized_name": "种子", + "name": "seed", + "type": "INT", + "widget": { + "name": "seed" + }, + "link": 54 + } + ], + "outputs": [ + { + "localized_name": "种子", + "name": "seed", + "type": "INT", + "links": [ + 52, + 53 + ] + } + ], + "properties": { + "Node name for S&R": "SeedNode", + "cnr_id": "comfy-core", + "ver": "0.31.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + 0, + "randomize" + ], + "widgets_values_named": { + "seed": 0, + "fixed": "randomize" + } + }, + "status": "type_match", + "metadata": { + "source_title": "SeedNode", + "api_meta": { + "title": "种子" + }, + "title_equal": false + } + }, + { + "api_id": "37:42", + "api_node": { + "inputs": { + "tile_size": 1536, + "overlap": 64, + "samples": [ + "37:9", + 0 + ], + "vae": [ + "37:7", + 0 + ] + }, + "class_type": "VAEDecodeAudioTiled", + "_meta": { + "title": "VAE解码音频(分块)" + } + }, + "fields": [ + { + "field": "tile_size", + "api_value": 1536, + "api_type": "int", + "named_evidence": { + "source_value": 1536, + "source_type": "int", + "status": "equal" + }, + "source_value": 1536, + "status": "named_equal" + }, + { + "field": "overlap", + "api_value": 64, + "api_type": "int", + "named_evidence": { + "source_value": 64, + "source_type": "int", + "status": "equal" + }, + "source_value": 64, + "status": "named_equal" + }, + { + "field": "samples", + "api_value": [ + "37:9", + 0 + ], + "api_type": "list", + "source_link": [ + 58, + 9, + 0, + 42, + 0, + "LATENT" + ], + "source_expected": [ + "37:9", + 0 + ], + "status": "link_equal" + }, + { + "field": "vae", + "api_value": [ + "37:7", + 0 + ], + "api_type": "list", + "source_link": [ + 59, + 7, + 0, + 42, + 1, + "VAE" + ], + "source_expected": [ + "37:7", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 42, + "type": "VAEDecodeAudioTiled", + "pos": [ + 2580, + 500 + ], + "size": [ + 280, + 110 + ], + "flags": {}, + "order": 9, + "mode": 0, + "inputs": [ + { + "localized_name": "样本", + "name": "samples", + "type": "LATENT", + "link": 58 + }, + { + "localized_name": "vae", + "name": "vae", + "type": "VAE", + "link": 59 + }, + { + "localized_name": "分块大小", + "name": "tile_size", + "type": "INT", + "widget": { + "name": "tile_size" + }, + "link": null + }, + { + "localized_name": "重叠", + "name": "overlap", + "type": "INT", + "widget": { + "name": "overlap" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "音频", + "name": "AUDIO", + "type": "AUDIO", + "links": [ + 61 + ] + } + ], + "properties": { + "Node name for S&R": "VAEDecodeAudioTiled", + "cnr_id": "comfy-core", + "ver": "0.31.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + 1536, + 64 + ], + "widgets_values_named": { + "tile_size": 1536, + "overlap": 64 + } + }, + "status": "type_match", + "metadata": { + "source_title": "VAEDecodeAudioTiled", + "api_meta": { + "title": "VAE解码音频(分块)" + }, + "title_equal": false + } + }, + { + "api_id": "37:43", + "api_node": { + "inputs": { + "switch": true, + "on_false": [ + "37:12", + 0 + ], + "on_true": [ + "37:42", + 0 + ] + }, + "class_type": "ComfySwitchNode", + "_meta": { + "title": "切换" + } + }, + "fields": [ + { + "field": "switch", + "api_value": true, + "api_type": "bool", + "named_evidence": { + "source_value": false, + "source_type": "bool", + "status": "different" + }, + "source_link": [ + 63, + -10, + 7, + 43, + 2, + "BOOLEAN" + ], + "source_expected": [ + "37:-10", + 7 + ], + "status": "link_difference" + }, + { + "field": "on_false", + "api_value": [ + "37:12", + 0 + ], + "api_type": "list", + "source_link": [ + 60, + 12, + 0, + 43, + 0, + "AUDIO" + ], + "source_expected": [ + "37:12", + 0 + ], + "status": "link_equal" + }, + { + "field": "on_true", + "api_value": [ + "37:42", + 0 + ], + "api_type": "list", + "source_link": [ + 61, + 42, + 0, + 43, + 1, + "AUDIO" + ], + "source_expected": [ + "37:42", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 43, + "type": "ComfySwitchNode", + "pos": [ + 2600, + 690 + ], + "size": [ + 270, + 110 + ], + "flags": {}, + "order": 10, + "mode": 0, + "inputs": [ + { + "localized_name": "为假时", + "name": "on_false", + "type": "AUDIO", + "link": 60 + }, + { + "localized_name": "为真时", + "name": "on_true", + "type": "AUDIO", + "link": 61 + }, + { + "localized_name": "切换", + "name": "switch", + "type": "BOOLEAN", + "widget": { + "name": "switch" + }, + "link": 63 + } + ], + "outputs": [ + { + "localized_name": "输出", + "name": "output", + "type": "AUDIO", + "links": [ + 62 + ] + } + ], + "properties": { + "Node name for S&R": "ComfySwitchNode", + "cnr_id": "comfy-core", + "ver": "0.31.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + false + ], + "widgets_values_named": { + "switch": false + } + }, + "status": "type_match", + "metadata": { + "source_title": "ComfySwitchNode", + "api_meta": { + "title": "切换" + }, + "title_equal": false + } + } + ], + "source_only_nodes": [ + { + "source_id": "39", + "node": { + "id": 39, + "type": "MarkdownNote", + "pos": [ + 1050, + 630 + ], + "size": [ + 520, + 660 + ], + "flags": {}, + "order": 0, + "mode": 0, + "inputs": [], + "outputs": [], + "title": "Note: MiniMax Music 3", + "properties": { + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "## MiniMax Music 3\n\nA music generation model by MiniMax that creates complete songs up to 5 minutes long with stable structure and high audio quality. Two inputs drive it: **Caption** (structured music description: style, mood, vocals, arrangement) and **Lyrics** (with `[intro]` `[verse]` `[chorus]` `[bridge]` `[outro]` tags controlling song structure).\n\n### Parameters\n\n| Parameter | Description |\n|---|---|\n| Caption | Music description. Write it in three sections — Global Metadata → Vocal Details → Arrangement. The more specific, the closer the result. |\n| Lyrics | Song lyrics + section tags. Tags are the only executable structural instructions; the lyric text itself only conveys mood. |\n| max_duration | Target song length in seconds (default 120 = 2 minutes; the model supports up to ~300 s / 5 minutes). Longer songs take more time and VRAM. |\n| seed | Random seed for the generation. Keep it fixed to reproduce the same song; change it to get a different take. |\n| Model files | diffusion model / text encoder / VAE.|\n| Tiled decode | Decode the audio VAE in overlapping tiles to drastically cut VRAM usage — helpful for long songs on low-VRAM GPUs. Slightly slower, with a small risk of seams at tile boundaries; turn it off on high-VRAM GPUs for the best quality. |\n\n### Prompt Writing Guide\n\nYou can find an official [Music Caption Rewriter Skill here](https://github.com/MiniMax-AI/MiniMax-Music3)" + ], + "widgets_values_named": { + "text": "## MiniMax Music 3\n\nA music generation model by MiniMax that creates complete songs up to 5 minutes long with stable structure and high audio quality. Two inputs drive it: **Caption** (structured music description: style, mood, vocals, arrangement) and **Lyrics** (with `[intro]` `[verse]` `[chorus]` `[bridge]` `[outro]` tags controlling song structure).\n\n### Parameters\n\n| Parameter | Description |\n|---|---|\n| Caption | Music description. Write it in three sections — Global Metadata → Vocal Details → Arrangement. The more specific, the closer the result. |\n| Lyrics | Song lyrics + section tags. Tags are the only executable structural instructions; the lyric text itself only conveys mood. |\n| max_duration | Target song length in seconds (default 120 = 2 minutes; the model supports up to ~300 s / 5 minutes). Longer songs take more time and VRAM. |\n| seed | Random seed for the generation. Keep it fixed to reproduce the same song; change it to get a different take. |\n| Model files | diffusion model / text encoder / VAE.|\n| Tiled decode | Decode the audio VAE in overlapping tiles to drastically cut VRAM usage — helpful for long songs on low-VRAM GPUs. Slightly slower, with a small risk of seams at tile boundaries; turn it off on high-VRAM GPUs for the best quality. |\n\n### Prompt Writing Guide\n\nYou can find an official [Music Caption Rewriter Skill here](https://github.com/MiniMax-AI/MiniMax-Music3)" + }, + "color": "#222", + "bgcolor": "#000" + } + }, + { + "source_id": "40", + "node": { + "id": 40, + "type": "MarkdownNote", + "pos": [ + 560, + 630 + ], + "size": [ + 440, + 680 + ], + "flags": {}, + "order": 1, + "mode": 0, + "inputs": [], + "outputs": [], + "properties": { + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "## Model Links\n\n**diffusion_models**\n- [minimax_music3_dit_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/MiniMax-Music-3/resolve/main/diffusion_models/minimax_music3_dit_int8_convrot.safetensors) # For low VRAM\n- [minimax_music3_dit_fp16.safetensors](https://huggingface.co/Comfy-Org/MiniMax-Music-3/resolve/main/diffusion_models/minimax_music3_dit_fp16.safetensors)\n\n**text_encoders**\n\n- [minimax_music3_text_encoder_pruned_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/MiniMax-Music-3/resolve/main/text_encoders/minimax_music3_text_encoder_pruned_int8_convrot.safetensors)\n\n**vae**\n\n- [minimax_music3_dav.safetensors](https://huggingface.co/Comfy-Org/MiniMax-Music-3/resolve/main/vae/minimax_music3_dav.safetensors)\n\n\n## Model Storage Location\n\n```\n📂 ComfyUI/\n├── 📂 models/\n│ ├── 📂 diffusion_models/\n│ │ ├── minimax_music3_dit_int8_convrot.safetensors # For low VRAM\n│ │ └── minimax_music3_dit_fp16.safetensors\n│ ├── 📂 text_encoders/\n│ │ └── minimax_music3_text_encoder_pruned_int8_convrot.safetensors\n│ └── 📂 vae/\n│ └── minimax_music3_dav.safetensors\n```\n\n## Report Issue\n\nNote: Please update ComfyUI first ([guide](https://docs.comfy.org/installation/update_comfyui)) and prepare required models. Desktop/Cloud updates follow stable releases, so some nightly-supported models may not be available yet.\n\n- Cannot run / runtime errors: [ComfyUI/issues](https://github.com/comfyanonymous/ComfyUI/issues)\n- UI / frontend issues: [ComfyUI_frontend/issues](https://github.com/Comfy-Org/ComfyUI_frontend/issues)\n- Workflow issues: [workflow_templates/issues](https://github.com/Comfy-Org/workflow_templates/issues)\n" + ], + "widgets_values_named": { + "text": "## Model Links\n\n**diffusion_models**\n- [minimax_music3_dit_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/MiniMax-Music-3/resolve/main/diffusion_models/minimax_music3_dit_int8_convrot.safetensors) # For low VRAM\n- [minimax_music3_dit_fp16.safetensors](https://huggingface.co/Comfy-Org/MiniMax-Music-3/resolve/main/diffusion_models/minimax_music3_dit_fp16.safetensors)\n\n**text_encoders**\n\n- [minimax_music3_text_encoder_pruned_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/MiniMax-Music-3/resolve/main/text_encoders/minimax_music3_text_encoder_pruned_int8_convrot.safetensors)\n\n**vae**\n\n- [minimax_music3_dav.safetensors](https://huggingface.co/Comfy-Org/MiniMax-Music-3/resolve/main/vae/minimax_music3_dav.safetensors)\n\n\n## Model Storage Location\n\n```\n📂 ComfyUI/\n├── 📂 models/\n│ ├── 📂 diffusion_models/\n│ │ ├── minimax_music3_dit_int8_convrot.safetensors # For low VRAM\n│ │ └── minimax_music3_dit_fp16.safetensors\n│ ├── 📂 text_encoders/\n│ │ └── minimax_music3_text_encoder_pruned_int8_convrot.safetensors\n│ └── 📂 vae/\n│ └── minimax_music3_dav.safetensors\n```\n\n## Report Issue\n\nNote: Please update ComfyUI first ([guide](https://docs.comfy.org/installation/update_comfyui)) and prepare required models. Desktop/Cloud updates follow stable releases, so some nightly-supported models may not be available yet.\n\n- Cannot run / runtime errors: [ComfyUI/issues](https://github.com/comfyanonymous/ComfyUI/issues)\n- UI / frontend issues: [ComfyUI_frontend/issues](https://github.com/Comfy-Org/ComfyUI_frontend/issues)\n- Workflow issues: [workflow_templates/issues](https://github.com/Comfy-Org/workflow_templates/issues)\n" + }, + "color": "#222", + "bgcolor": "#000" + } + }, + { + "source_id": "37", + "node": { + "id": 37, + "type": "ac99f841-a3de-4329-9564-953b81cf9e16", + "pos": [ + 1620, + 630 + ], + "size": [ + 490, + 540 + ], + "flags": {}, + "order": 2, + "mode": 0, + "inputs": [ + { + "name": "caption", + "type": "STRING", + "widget": { + "name": "caption" + }, + "link": null + }, + { + "name": "lyrics", + "type": "STRING", + "widget": { + "name": "lyrics" + }, + "link": null + }, + { + "name": "max_duration", + "type": "FLOAT", + "widget": { + "name": "max_duration" + }, + "link": null + }, + { + "name": "seed", + "type": "INT", + "widget": { + "name": "seed" + }, + "link": null + }, + { + "name": "unet_name", + "type": "COMBO", + "widget": { + "name": "unet_name" + }, + "link": null + }, + { + "name": "clip_name", + "type": "COMBO", + "widget": { + "name": "clip_name" + }, + "link": null + }, + { + "name": "vae_name", + "type": "COMBO", + "widget": { + "name": "vae_name" + }, + "link": null + }, + { + "label": "tiled_decode", + "name": "switch", + "type": "BOOLEAN", + "widget": { + "name": "switch" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "AUDIO", + "name": "AUDIO", + "type": "AUDIO", + "links": [ + 42 + ] + } + ], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.31.0", + "previewExposures": [], + "ue_properties": { + "widget_ue_connectable": { + "switch": true + }, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "Global Metadata: Lo-fi hip-hop, chillhop. 78 BPM, D flat major, major scale with jazzy extensions. Laid-back and dreamy throughout, a gentle warm drift with a subtle late-night glow that deepens in the middle and dissolves softly at the end. Studying, raining-outside, headphones-on late-night listening. Bedroom production: muddy warm texture, heavy vinyl crackle, tape hiss and wow-flutter pitch wobble, low-passed dusty mix, soft-clipped drums, everything slightly detuned and cozy.\n\nVocal Details: Soft androgynous vocal, hushed half-sung half-spoken delivery, sitting low in the mix like another instrument, lazy behind-the-beat phrasing, gentle breathy timbre. Sparse murmured double-tracked harmonies, occasional wordless \"mmm\" and \"ooh\" hums drenched in tape delay and warm spring reverb. Long stretches with no vocals at all.\n\nArrangement: Dusty boom-bap drums with a soft thumping kick, cracked snare with lazy swing, brushed hi-hats, low round sub bass. Warm Rhodes piano chords with slow chorus wobble as the harmonic bed, mellow jazzy guitar licks answering the vocal lines, constant vinyl crackle as texture. Intro: rain and vinyl noise, solo Rhodes chords fading in, drums slipping in halfway. Verses: minimal — drums, bass, Rhodes, soft guitar fills between lines. Instrumental sections: guitar and Rhodes trade relaxed jazzy phrases over the beat, occasional muted trumpet ghost notes far in the background. Bridge: drums drop away to rain, crackle, and floating detuned Rhodes, then the beat eases back in. Outro: elements fade one by one until only vinyl crackle and a last unresolved Rhodes chord remain.", + "[Intro]\nMmm...\n(rain on the window)\nOoh...\n\n[Verse]\nMidnight and the canvas glows\nDragging little wires where the current flows\nType a quiet dream, let the sampler drift\nNoise into a picture, like the fog just lifts\nTwenty slow steps, I'm in no hurry now\nLatents turning colors and I don't know how\nEvery render's like a polaroid I found\nSoft focus memories, no sound\n\n[Instrumental]\n\n[Verse]\nQueue another frame, let the motion breathe\nPictures start to move like the falling leaves\nVideo drifting by at twenty-four\nLittle animations on my bedroom floor\nSeed after seed like the rain outside\nSome of them are keepers, some I let slide\nSave the ones that feel like a Sunday slow\nNode to node to node... and off we go\n\n[Chorus]\nMmm... let it render on\n(take your time, take your time)\nOoh... by the morning it'll all be done\n(one more queue, one more try)\n\n[Instrumental]\n\n[Bridge]\nRain keeps drawing pictures on the glass...\nMy machine keeps dreaming...\nNeither of us fast...\n\n[Chorus]\nMmm... let it render on\n(take your time, take your time)\nOoh... by the morning it'll all be done\n(one more queue, one more try)\n\n[Instrumental]\n\n[Outro]\nMmm...\n(node to node)\nOoh... goodnight", + 60, + 617970236276137, + "MiniMax\\minimax_music3_dit_fp16.safetensors", + "MiniMax\\minimax_music3_text_encoder_pruned_int8_convrot.safetensors", + "MiniMax\\minimax_music3_dav.safetensors", + true + ], + "widgets_values_named": { + "caption": "Global Metadata: Lo-fi hip-hop, chillhop. 78 BPM, D flat major, major scale with jazzy extensions. Laid-back and dreamy throughout, a gentle warm drift with a subtle late-night glow that deepens in the middle and dissolves softly at the end. Studying, raining-outside, headphones-on late-night listening. Bedroom production: muddy warm texture, heavy vinyl crackle, tape hiss and wow-flutter pitch wobble, low-passed dusty mix, soft-clipped drums, everything slightly detuned and cozy.\n\nVocal Details: Soft androgynous vocal, hushed half-sung half-spoken delivery, sitting low in the mix like another instrument, lazy behind-the-beat phrasing, gentle breathy timbre. Sparse murmured double-tracked harmonies, occasional wordless \"mmm\" and \"ooh\" hums drenched in tape delay and warm spring reverb. Long stretches with no vocals at all.\n\nArrangement: Dusty boom-bap drums with a soft thumping kick, cracked snare with lazy swing, brushed hi-hats, low round sub bass. Warm Rhodes piano chords with slow chorus wobble as the harmonic bed, mellow jazzy guitar licks answering the vocal lines, constant vinyl crackle as texture. Intro: rain and vinyl noise, solo Rhodes chords fading in, drums slipping in halfway. Verses: minimal — drums, bass, Rhodes, soft guitar fills between lines. Instrumental sections: guitar and Rhodes trade relaxed jazzy phrases over the beat, occasional muted trumpet ghost notes far in the background. Bridge: drums drop away to rain, crackle, and floating detuned Rhodes, then the beat eases back in. Outro: elements fade one by one until only vinyl crackle and a last unresolved Rhodes chord remain.", + "lyrics": "[Intro]\nMmm...\n(rain on the window)\nOoh...\n\n[Verse]\nMidnight and the canvas glows\nDragging little wires where the current flows\nType a quiet dream, let the sampler drift\nNoise into a picture, like the fog just lifts\nTwenty slow steps, I'm in no hurry now\nLatents turning colors and I don't know how\nEvery render's like a polaroid I found\nSoft focus memories, no sound\n\n[Instrumental]\n\n[Verse]\nQueue another frame, let the motion breathe\nPictures start to move like the falling leaves\nVideo drifting by at twenty-four\nLittle animations on my bedroom floor\nSeed after seed like the rain outside\nSome of them are keepers, some I let slide\nSave the ones that feel like a Sunday slow\nNode to node to node... and off we go\n\n[Chorus]\nMmm... let it render on\n(take your time, take your time)\nOoh... by the morning it'll all be done\n(one more queue, one more try)\n\n[Instrumental]\n\n[Bridge]\nRain keeps drawing pictures on the glass...\nMy machine keeps dreaming...\nNeither of us fast...\n\n[Chorus]\nMmm... let it render on\n(take your time, take your time)\nOoh... by the morning it'll all be done\n(one more queue, one more try)\n\n[Instrumental]\n\n[Outro]\nMmm...\n(node to node)\nOoh... goodnight", + "max_duration": 60, + "seed": 617970236276137, + "unet_name": "MiniMax\\minimax_music3_dit_fp16.safetensors", + "clip_name": "MiniMax\\minimax_music3_text_encoder_pruned_int8_convrot.safetensors", + "vae_name": "MiniMax\\minimax_music3_dav.safetensors", + "switch": true + } + } + } + ], + "unconfirmed": [ + "虚拟节点、子图边界、开关与循环展开语义需独立核查", + "未有可信节点定义;单位、范围、环境和实际产物未验证" + ] + } +} diff --git a/docs/hel-892/sixth-evidence/04.json b/docs/hel-892/sixth-evidence/04.json new file mode 100644 index 00000000..023a9db2 --- /dev/null +++ b/docs/hel-892/sixth-evidence/04.json @@ -0,0 +1,3223 @@ +{ + "binding": { + "id": "04", + "source_path": "example_ComfyUI_workflows/3-图像生成/Qwen-Image-2.1/Qwen_image2.1_文生图_提示词增强.json", + "source_sha": "791b23c19730a5a88dbf19247bfae5c814df9741", + "export_file": "04.api.json", + "source_sha256": "dfe10917d101bf1a8a8c20750f4360690ecd478e9eb33d16a59ea0c6f12c086c", + "export_sha256": "2066b07b2ca5390ba3e795df6b43934dab83865b12870b6ff26d8c0ef24f091e", + "original_filename": "Qwen_image2.1_文生图_提示词增强.json", + "provider": "白栖知", + "attachment_id": "01a187e2-3666-7bc7-9101-98fe3dab8070", + "received_at": "2026-10-29T13:38:38Z", + "received_at_basis": "主卡附件评论创建时间;不是导出时间", + "source_comment_id": "01a187e2-4f6e-761e-9412-43042e094daa", + "exported_at": null, + "frontend_version": null, + "node_versions": null, + "provenance": "user_supplied_api_export_environment_unknown", + "source_export_equivalence_verified": false, + "instruction": "文生图:13输出分辨率,483.value提供提示词,459子图展开;459:471为TextGenerate,459:474开关连接提示词分支,459:452编码后经459:458采样及459:457解码送461保存。具名采样字段和嵌套配置原值来自API,不由位置控件推断;开关行为与模型版本待核查。" + }, + "adopted": true, + "method": "supplied_api_companion", + "general_conversion_verified": false, + "execution_verified": false, + "comparison": { + "equivalent": false, + "scope": "逐节点/字段保留API值和类型;直接连线比较ID及槽;位置控件不按顺序猜名", + "nodes": [ + { + "api_id": "13", + "api_node": { + "inputs": { + "aspect_ratio": "3:4 (Portrait Standard)", + "megapixels": 1.5, + "multiple": 8 + }, + "class_type": "ResolutionSelector", + "_meta": { + "title": "分辨率选择器" + } + }, + "fields": [ + { + "field": "aspect_ratio", + "api_value": "3:4 (Portrait Standard)", + "api_type": "str", + "named_evidence": { + "source_value": "3:4 (Portrait Standard)", + "source_type": "str", + "status": "equal" + }, + "source_value": "3:4 (Portrait Standard)", + "status": "named_equal" + }, + { + "field": "megapixels", + "api_value": 1.5, + "api_type": "float", + "named_evidence": { + "source_value": 1.5, + "source_type": "float", + "status": "equal" + }, + "source_value": 1.5, + "status": "named_equal" + }, + { + "field": "multiple", + "api_value": 8, + "api_type": "int", + "named_evidence": { + "source_value": 8, + "source_type": "int", + "status": "equal" + }, + "source_value": 8, + "status": "named_equal" + } + ], + "source_node": { + "id": 13, + "type": "ResolutionSelector", + "pos": [ + 510, + 2820 + ], + "size": [ + 310, + 150 + ], + "flags": {}, + "order": 0, + "mode": 0, + "showAdvanced": false, + "inputs": [ + { + "localized_name": "宽高比", + "name": "aspect_ratio", + "type": "COMBO", + "widget": { + "name": "aspect_ratio" + }, + "link": null + }, + { + "localized_name": "百万像素", + "name": "megapixels", + "type": "FLOAT", + "widget": { + "name": "megapixels" + }, + "link": null + }, + { + "localized_name": "倍数", + "name": "multiple", + "type": "INT", + "widget": { + "name": "multiple" + }, + "link": null + }, + { + "localized_name": "preview", + "name": "preview", + "shape": 7, + "type": "RESOLUTION_PREVIEW", + "widget": { + "name": "preview" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "宽度", + "name": "width", + "type": "INT", + "links": [ + 670 + ] + }, + { + "localized_name": "高度", + "name": "height", + "type": "INT", + "links": [ + 671 + ] + } + ], + "properties": { + "Node name for S&R": "ResolutionSelector", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "3:4 (Portrait Standard)", + 1.5, + 8 + ], + "widgets_values_named": { + "aspect_ratio": "3:4 (Portrait Standard)", + "megapixels": 1.5, + "multiple": 8 + } + }, + "status": "type_match", + "metadata": { + "source_title": "ResolutionSelector", + "api_meta": { + "title": "分辨率选择器" + }, + "title_equal": false + } + }, + { + "api_id": "461", + "api_node": { + "inputs": { + "filename_prefix": "image/Qwen_image_2.1", + "format": "png", + "format.bit_depth": "8-bit", + "format.input_color_space": "sRGB", + "images": [ + "459:457", + 0 + ] + }, + "class_type": "SaveImageAdvanced", + "_meta": { + "title": "保存图像(高级)" + } + }, + "fields": [ + { + "field": "filename_prefix", + "api_value": "image/Qwen_image_2.1", + "api_type": "str", + "named_evidence": { + "source_value": "image/Qwen_image_2.1", + "source_type": "str", + "status": "equal" + }, + "source_value": "image/Qwen_image_2.1", + "status": "named_equal" + }, + { + "field": "format", + "api_value": "png", + "api_type": "str", + "named_evidence": { + "source_value": "png", + "source_type": "str", + "status": "equal" + }, + "source_value": "png", + "status": "named_equal" + }, + { + "field": "format.bit_depth", + "api_value": "8-bit", + "api_type": "str", + "named_evidence": { + "source_value": "8-bit", + "source_type": "str", + "status": "equal" + }, + "source_value": "8-bit", + "status": "named_equal" + }, + { + "field": "format.input_color_space", + "api_value": "sRGB", + "api_type": "str", + "named_evidence": { + "source_value": "sRGB", + "source_type": "str", + "status": "equal" + }, + "source_value": "sRGB", + "status": "named_equal" + }, + { + "field": "images", + "api_value": [ + "459:457", + 0 + ], + "api_type": "list", + "source_link": [ + 672, + 459, + 0, + 461, + 0, + "IMAGE" + ], + "source_expected": [ + "459", + 0 + ], + "status": "link_difference" + } + ], + "source_node": { + "id": 461, + "type": "SaveImageAdvanced", + "pos": [ + 1400, + 2460 + ], + "size": [ + 850, + 610 + ], + "flags": { + "pinned": true + }, + "order": 5, + "mode": 0, + "inputs": [ + { + "localized_name": "图像", + "name": "images", + "type": "IMAGE", + "link": 672 + }, + { + "localized_name": "文件名前缀", + "name": "filename_prefix", + "type": "STRING", + "widget": { + "name": "filename_prefix" + }, + "link": null + }, + { + "localized_name": "格式", + "name": "format", + "type": "COMFY_DYNAMICCOMBO_V3", + "widget": { + "name": "format" + }, + "link": null + }, + { + "localized_name": "位深度", + "name": "format.bit_depth", + "type": "COMBO", + "widget": { + "name": "format.bit_depth" + }, + "link": null + }, + { + "localized_name": "输入色彩空间", + "name": "format.input_color_space", + "type": "COMBO", + "widget": { + "name": "format.input_color_space" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "images", + "name": "images", + "type": "IMAGE", + "links": [] + } + ], + "properties": { + "Node name for S&R": "SaveImageAdvanced", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "image/Qwen_image_2.1", + "png", + "8-bit", + "sRGB" + ], + "widgets_values_named": { + "filename_prefix": "image/Qwen_image_2.1", + "format": "png", + "format.bit_depth": "8-bit", + "format.input_color_space": "sRGB" + } + }, + "status": "type_match", + "metadata": { + "source_title": "SaveImageAdvanced", + "api_meta": { + "title": "保存图像(高级)" + }, + "title_equal": false + } + }, + { + "api_id": "483", + "api_node": { + "inputs": { + "value": "角色设定卡,国风清雅,宣纸米白背景,淡青灰晕染。左侧为22岁中国女性面部特写,温婉沉静,鹅蛋脸,远山黛眉,杏眼深茶褐,淡樱粉唇,乌黑长发半绾青玉簪,鬓角碎发,淡水珍珠耳钉。右侧为人物三视图:正面、侧面、背面,全身站姿,7.5头身,双手交叠身前,身姿端庄。身穿雨过天青真丝提花缎改良旗袍,立领,短袖,收腰,A字裙摆,长至小腿中下部,两侧开衩至膝上,手绘水墨竹纹,竹节深青勾勒,竹叶淡青晕染,竹节形深青盘扣5枚,领口袖口开衩深青细滚边,同色真丝内衬,米白真丝绣花鞋低跟竹叶刺绣。右下角服装细节放大:领口竹节盘扣、水墨竹纹、袖口滚边、开衩暗扣、青玉簪、珍珠耳钉、绣花鞋、提花暗纹。底部色卡:雨过天青、深青、月白、玉色、墨青、珍珠白。画面留出文字框区域:左上标题框、左侧角色信息框、右侧三视图标注框、下方服装说明框、底部设计说明框。柔和自然光,工笔国风,超高细节,角色设计参考图,干净排版。" + }, + "class_type": "PrimitiveStringMultiline", + "_meta": { + "title": "字符串(多行)" + } + }, + "fields": [ + { + "field": "value", + "api_value": "角色设定卡,国风清雅,宣纸米白背景,淡青灰晕染。左侧为22岁中国女性面部特写,温婉沉静,鹅蛋脸,远山黛眉,杏眼深茶褐,淡樱粉唇,乌黑长发半绾青玉簪,鬓角碎发,淡水珍珠耳钉。右侧为人物三视图:正面、侧面、背面,全身站姿,7.5头身,双手交叠身前,身姿端庄。身穿雨过天青真丝提花缎改良旗袍,立领,短袖,收腰,A字裙摆,长至小腿中下部,两侧开衩至膝上,手绘水墨竹纹,竹节深青勾勒,竹叶淡青晕染,竹节形深青盘扣5枚,领口袖口开衩深青细滚边,同色真丝内衬,米白真丝绣花鞋低跟竹叶刺绣。右下角服装细节放大:领口竹节盘扣、水墨竹纹、袖口滚边、开衩暗扣、青玉簪、珍珠耳钉、绣花鞋、提花暗纹。底部色卡:雨过天青、深青、月白、玉色、墨青、珍珠白。画面留出文字框区域:左上标题框、左侧角色信息框、右侧三视图标注框、下方服装说明框、底部设计说明框。柔和自然光,工笔国风,超高细节,角色设计参考图,干净排版。", + "api_type": "str", + "named_evidence": { + "source_value": "角色设定卡,国风清雅,宣纸米白背景,淡青灰晕染。左侧为22岁中国女性面部特写,温婉沉静,鹅蛋脸,远山黛眉,杏眼深茶褐,淡樱粉唇,乌黑长发半绾青玉簪,鬓角碎发,淡水珍珠耳钉。右侧为人物三视图:正面、侧面、背面,全身站姿,7.5头身,双手交叠身前,身姿端庄。身穿雨过天青真丝提花缎改良旗袍,立领,短袖,收腰,A字裙摆,长至小腿中下部,两侧开衩至膝上,手绘水墨竹纹,竹节深青勾勒,竹叶淡青晕染,竹节形深青盘扣5枚,领口袖口开衩深青细滚边,同色真丝内衬,米白真丝绣花鞋低跟竹叶刺绣。右下角服装细节放大:领口竹节盘扣、水墨竹纹、袖口滚边、开衩暗扣、青玉簪、珍珠耳钉、绣花鞋、提花暗纹。底部色卡:雨过天青、深青、月白、玉色、墨青、珍珠白。画面留出文字框区域:左上标题框、左侧角色信息框、右侧三视图标注框、下方服装说明框、底部设计说明框。柔和自然光,工笔国风,超高细节,角色设计参考图,干净排版。", + "source_type": "str", + "status": "equal" + }, + "source_value": "角色设定卡,国风清雅,宣纸米白背景,淡青灰晕染。左侧为22岁中国女性面部特写,温婉沉静,鹅蛋脸,远山黛眉,杏眼深茶褐,淡樱粉唇,乌黑长发半绾青玉簪,鬓角碎发,淡水珍珠耳钉。右侧为人物三视图:正面、侧面、背面,全身站姿,7.5头身,双手交叠身前,身姿端庄。身穿雨过天青真丝提花缎改良旗袍,立领,短袖,收腰,A字裙摆,长至小腿中下部,两侧开衩至膝上,手绘水墨竹纹,竹节深青勾勒,竹叶淡青晕染,竹节形深青盘扣5枚,领口袖口开衩深青细滚边,同色真丝内衬,米白真丝绣花鞋低跟竹叶刺绣。右下角服装细节放大:领口竹节盘扣、水墨竹纹、袖口滚边、开衩暗扣、青玉簪、珍珠耳钉、绣花鞋、提花暗纹。底部色卡:雨过天青、深青、月白、玉色、墨青、珍珠白。画面留出文字框区域:左上标题框、左侧角色信息框、右侧三视图标注框、下方服装说明框、底部设计说明框。柔和自然光,工笔国风,超高细节,角色设计参考图,干净排版。", + "status": "named_equal" + } + ], + "source_node": { + "id": 483, + "type": "PrimitiveStringMultiline", + "pos": [ + 490, + 2480 + ], + "size": [ + 370, + 280 + ], + "flags": {}, + "order": 3, + "mode": 0, + "inputs": [ + { + "localized_name": "值", + "name": "value", + "type": "STRING", + "widget": { + "name": "value" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "字符串", + "name": "STRING", + "type": "STRING", + "links": [ + 713 + ] + } + ], + "properties": { + "Node name for S&R": "PrimitiveStringMultiline", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "ue_properties": { + "widget_ue_connectable": {}, + "input_ue_unconnectable": {}, + "version": "7.8" + } + }, + "widgets_values": [ + "角色设定卡,国风清雅,宣纸米白背景,淡青灰晕染。左侧为22岁中国女性面部特写,温婉沉静,鹅蛋脸,远山黛眉,杏眼深茶褐,淡樱粉唇,乌黑长发半绾青玉簪,鬓角碎发,淡水珍珠耳钉。右侧为人物三视图:正面、侧面、背面,全身站姿,7.5头身,双手交叠身前,身姿端庄。身穿雨过天青真丝提花缎改良旗袍,立领,短袖,收腰,A字裙摆,长至小腿中下部,两侧开衩至膝上,手绘水墨竹纹,竹节深青勾勒,竹叶淡青晕染,竹节形深青盘扣5枚,领口袖口开衩深青细滚边,同色真丝内衬,米白真丝绣花鞋低跟竹叶刺绣。右下角服装细节放大:领口竹节盘扣、水墨竹纹、袖口滚边、开衩暗扣、青玉簪、珍珠耳钉、绣花鞋、提花暗纹。底部色卡:雨过天青、深青、月白、玉色、墨青、珍珠白。画面留出文字框区域:左上标题框、左侧角色信息框、右侧三视图标注框、下方服装说明框、底部设计说明框。柔和自然光,工笔国风,超高细节,角色设计参考图,干净排版。" + ], + "widgets_values_named": { + "value": "角色设定卡,国风清雅,宣纸米白背景,淡青灰晕染。左侧为22岁中国女性面部特写,温婉沉静,鹅蛋脸,远山黛眉,杏眼深茶褐,淡樱粉唇,乌黑长发半绾青玉簪,鬓角碎发,淡水珍珠耳钉。右侧为人物三视图:正面、侧面、背面,全身站姿,7.5头身,双手交叠身前,身姿端庄。身穿雨过天青真丝提花缎改良旗袍,立领,短袖,收腰,A字裙摆,长至小腿中下部,两侧开衩至膝上,手绘水墨竹纹,竹节深青勾勒,竹叶淡青晕染,竹节形深青盘扣5枚,领口袖口开衩深青细滚边,同色真丝内衬,米白真丝绣花鞋低跟竹叶刺绣。右下角服装细节放大:领口竹节盘扣、水墨竹纹、袖口滚边、开衩暗扣、青玉簪、珍珠耳钉、绣花鞋、提花暗纹。底部色卡:雨过天青、深青、月白、玉色、墨青、珍珠白。画面留出文字框区域:左上标题框、左侧角色信息框、右侧三视图标注框、下方服装说明框、底部设计说明框。柔和自然光,工笔国风,超高细节,角色设计参考图,干净排版。" + } + }, + "status": "type_match", + "metadata": { + "source_title": "PrimitiveStringMultiline", + "api_meta": { + "title": "字符串(多行)" + }, + "title_equal": false + } + }, + { + "api_id": "459:451", + "api_node": { + "inputs": { + "unet_name": "Qwen-Image/qwen_image_2.1_bf16.safetensors", + "weight_dtype": "default" + }, + "class_type": "UNETLoader", + "_meta": { + "title": "UNet加载器" + } + }, + "fields": [ + { + "field": "unet_name", + "api_value": "Qwen-Image/qwen_image_2.1_bf16.safetensors", + "api_type": "str", + "named_evidence": { + "source_value": "qwen_image_2.1_int8_convrot.safetensors", + "source_type": "str", + "status": "different" + }, + "source_link": [ + 663, + -10, + 11, + 451, + 0, + "COMBO" + ], + "source_expected": [ + "459:-10", + 11 + ], + "status": "link_difference" + }, + { + "field": "weight_dtype", + "api_value": "default", + "api_type": "str", + "named_evidence": { + "source_value": "default", + "source_type": "str", + "status": "equal" + }, + "source_value": "default", + "status": "named_equal" + } + ], + "source_node": { + "id": 451, + "type": "UNETLoader", + "pos": [ + -190, + 410 + ], + "size": [ + 510, + 90 + ], + "flags": {}, + "order": 1, + "mode": 0, + "inputs": [ + { + "localized_name": "UNet名称", + "name": "unet_name", + "type": "COMBO", + "widget": { + "name": "unet_name" + }, + "link": 663 + }, + { + "localized_name": "数据类型", + "name": "weight_dtype", + "type": "COMBO", + "widget": { + "name": "weight_dtype" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "模型", + "name": "MODEL", + "type": "MODEL", + "links": [ + 701 + ] + } + ], + "properties": { + "Node name for S&R": "UNETLoader", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "models": [ + { + "name": "qwen_image_2.1_int8_convrot.safetensors", + "url": "https://huggingface.co/Comfy-Org/Qwen-Image-2.1/resolve/main/diffusion_models/qwen_image_2.1_int8_convrot.safetensors", + "directory": "diffusion_models" + } + ], + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "qwen_image_2.1_int8_convrot.safetensors", + "default" + ], + "widgets_values_named": { + "unet_name": "qwen_image_2.1_int8_convrot.safetensors", + "weight_dtype": "default" + } + }, + "status": "type_match", + "metadata": { + "source_title": "UNETLoader", + "api_meta": { + "title": "UNet加载器" + }, + "title_equal": false + } + }, + { + "api_id": "459:452", + "api_node": { + "inputs": { + "prompt": [ + "459:472", + 0 + ], + "negative_prompt": "", + "resolution": 1024, + "clip": [ + "459:453", + 0 + ] + }, + "class_type": "TextEncodeQwenImage21", + "_meta": { + "title": "文本编码 Qwen Image 2.1" + } + }, + "fields": [ + { + "field": "prompt", + "api_value": [ + "459:472", + 0 + ], + "api_type": "list", + "named_evidence": { + "source_value": "", + "source_type": "str", + "status": "different" + }, + "source_link": [ + 687, + 472, + 0, + 452, + 3, + "STRING" + ], + "source_expected": [ + "459:472", + 0 + ], + "status": "link_equal" + }, + { + "field": "negative_prompt", + "api_value": "", + "api_type": "str", + "named_evidence": { + "source_value": "", + "source_type": "str", + "status": "equal" + }, + "source_link": [ + 657, + -10, + 3, + 452, + 4, + "STRING" + ], + "source_expected": [ + "459:-10", + 3 + ], + "status": "link_difference" + }, + { + "field": "resolution", + "api_value": 1024, + "api_type": "int", + "named_evidence": { + "source_value": 1024, + "source_type": "int", + "status": "equal" + }, + "source_value": 1024, + "status": "named_equal" + }, + { + "field": "clip", + "api_value": [ + "459:453", + 0 + ], + "api_type": "list", + "source_link": [ + 647, + 453, + 0, + 452, + 0, + "CLIP" + ], + "source_expected": [ + "459:453", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 452, + "type": "TextEncodeQwenImage21", + "pos": [ + 440, + 410 + ], + "size": [ + 530, + 680 + ], + "flags": {}, + "order": 2, + "mode": 0, + "inputs": [ + { + "localized_name": "clip", + "name": "clip", + "type": "CLIP", + "link": 647 + }, + { + "localized_name": "image_1", + "name": "images.image_1", + "shape": 7, + "type": "IMAGE", + "link": null + }, + { + "localized_name": "vae", + "name": "vae", + "shape": 7, + "type": "VAE", + "link": null + }, + { + "localized_name": "prompt", + "name": "prompt", + "type": "STRING", + "widget": { + "name": "prompt" + }, + "link": 687 + }, + { + "localized_name": "negative_prompt", + "name": "negative_prompt", + "type": "STRING", + "widget": { + "name": "negative_prompt" + }, + "link": 657 + }, + { + "localized_name": "resolution", + "name": "resolution", + "type": "INT", + "widget": { + "name": "resolution" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "positive", + "name": "positive", + "type": "CONDITIONING", + "links": [ + 653 + ] + }, + { + "localized_name": "negative", + "name": "negative", + "type": "CONDITIONING", + "links": [ + 655 + ] + }, + { + "localized_name": "latent", + "name": "latent", + "type": "LATENT", + "links": null + } + ], + "properties": { + "Node name for S&R": "TextEncodeQwenImage21", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "", + "", + 1024 + ], + "widgets_values_named": { + "prompt": "", + "negative_prompt": "", + "resolution": 1024 + } + }, + "status": "type_match", + "metadata": { + "source_title": "TextEncodeQwenImage21", + "api_meta": { + "title": "文本编码 Qwen Image 2.1" + }, + "title_equal": false + } + }, + { + "api_id": "459:453", + "api_node": { + "inputs": { + "clip_name": "qwen3vl_8b_int8_convrot.safetensors", + "type": "qwen_image", + "device": "default" + }, + "class_type": "CLIPLoader", + "_meta": { + "title": "加载CLIP" + } + }, + "fields": [ + { + "field": "clip_name", + "api_value": "qwen3vl_8b_int8_convrot.safetensors", + "api_type": "str", + "named_evidence": { + "source_value": "qwen3vl_8b_int8_convrot.safetensors", + "source_type": "str", + "status": "equal" + }, + "source_link": [ + 664, + -10, + 12, + 453, + 0, + "COMBO" + ], + "source_expected": [ + "459:-10", + 12 + ], + "status": "link_difference" + }, + { + "field": "type", + "api_value": "qwen_image", + "api_type": "str", + "named_evidence": { + "source_value": "qwen_image", + "source_type": "str", + "status": "equal" + }, + "source_value": "qwen_image", + "status": "named_equal" + }, + { + "field": "device", + "api_value": "default", + "api_type": "str", + "named_evidence": { + "source_value": "default", + "source_type": "str", + "status": "equal" + }, + "source_value": "default", + "status": "named_equal" + } + ], + "source_node": { + "id": 453, + "type": "CLIPLoader", + "pos": [ + -190, + 570 + ], + "size": [ + 510, + 120 + ], + "flags": {}, + "order": 3, + "mode": 0, + "inputs": [ + { + "localized_name": "CLIP名称", + "name": "clip_name", + "type": "COMBO", + "widget": { + "name": "clip_name" + }, + "link": 664 + }, + { + "localized_name": "类型", + "name": "type", + "type": "COMBO", + "widget": { + "name": "type" + }, + "link": null + }, + { + "localized_name": "设备", + "name": "device", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "device" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "CLIP", + "name": "CLIP", + "type": "CLIP", + "links": [ + 647 + ] + } + ], + "properties": { + "Node name for S&R": "CLIPLoader", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "models": [ + { + "name": "qwen3vl_8b_int8_convrot.safetensors", + "url": "https://huggingface.co/Comfy-Org/Qwen-Image-2.1/resolve/main/text_encoders/qwen3vl_8b_int8_convrot.safetensors", + "directory": "text_encoders" + } + ], + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "qwen3vl_8b_int8_convrot.safetensors", + "qwen_image", + "default" + ], + "widgets_values_named": { + "clip_name": "qwen3vl_8b_int8_convrot.safetensors", + "type": "qwen_image", + "device": "default" + } + }, + "status": "type_match", + "metadata": { + "source_title": "CLIPLoader", + "api_meta": { + "title": "加载CLIP" + }, + "title_equal": false + } + }, + { + "api_id": "459:454", + "api_node": { + "inputs": { + "vae_name": "qwen_image_2.1_vae_bf16.safetensors" + }, + "class_type": "VAELoader", + "_meta": { + "title": "加载VAE" + } + }, + "fields": [ + { + "field": "vae_name", + "api_value": "qwen_image_2.1_vae_bf16.safetensors", + "api_type": "str", + "named_evidence": { + "source_value": "qwen_image_2.1_vae_bf16.safetensors", + "source_type": "str", + "status": "equal" + }, + "source_link": [ + 665, + -10, + 13, + 454, + 0, + "COMBO" + ], + "source_expected": [ + "459:-10", + 13 + ], + "status": "link_difference" + } + ], + "source_node": { + "id": 454, + "type": "VAELoader", + "pos": [ + -190, + 770 + ], + "size": [ + 510, + 70 + ], + "flags": {}, + "order": 4, + "mode": 0, + "inputs": [ + { + "localized_name": "vae名称", + "name": "vae_name", + "type": "COMBO", + "widget": { + "name": "vae_name" + }, + "link": 665 + } + ], + "outputs": [ + { + "localized_name": "VAE", + "name": "VAE", + "type": "VAE", + "links": [ + 649 + ] + } + ], + "properties": { + "Node name for S&R": "VAELoader", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "models": [ + { + "name": "qwen_image_2.1_vae_bf16.safetensors", + "url": "https://huggingface.co/Comfy-Org/Qwen-Image-2.1/resolve/main/vae/qwen_image_2.1_vae_bf16.safetensors", + "directory": "vae" + } + ], + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "qwen_image_2.1_vae_bf16.safetensors" + ], + "widgets_values_named": { + "vae_name": "qwen_image_2.1_vae_bf16.safetensors" + } + }, + "status": "type_match", + "metadata": { + "source_title": "VAELoader", + "api_meta": { + "title": "加载VAE" + }, + "title_equal": false + } + }, + { + "api_id": "459:456", + "api_node": { + "inputs": { + "width": [ + "13", + 0 + ], + "height": [ + "13", + 1 + ], + "batch_size": 1 + }, + "class_type": "EmptyLatentImage", + "_meta": { + "title": "空Latent图像" + } + }, + "fields": [ + { + "field": "width", + "api_value": [ + "13", + 0 + ], + "api_type": "list", + "named_evidence": { + "source_value": 1024, + "source_type": "int", + "status": "different" + }, + "source_link": [ + 660, + -10, + 6, + 456, + 0, + "INT" + ], + "source_expected": [ + "459:-10", + 6 + ], + "status": "link_difference" + }, + { + "field": "height", + "api_value": [ + "13", + 1 + ], + "api_type": "list", + "named_evidence": { + "source_value": 1024, + "source_type": "int", + "status": "different" + }, + "source_link": [ + 661, + -10, + 7, + 456, + 1, + "INT" + ], + "source_expected": [ + "459:-10", + 7 + ], + "status": "link_difference" + }, + { + "field": "batch_size", + "api_value": 1, + "api_type": "int", + "named_evidence": { + "source_value": 1, + "source_type": "int", + "status": "equal" + }, + "source_value": 1, + "status": "named_equal" + } + ], + "source_node": { + "id": 456, + "type": "EmptyLatentImage", + "pos": [ + -70, + 1010 + ], + "size": [ + 300, + 120 + ], + "flags": {}, + "order": 5, + "mode": 0, + "inputs": [ + { + "localized_name": "宽度", + "name": "width", + "type": "INT", + "widget": { + "name": "width" + }, + "link": 660 + }, + { + "localized_name": "高度", + "name": "height", + "type": "INT", + "widget": { + "name": "height" + }, + "link": 661 + }, + { + "localized_name": "批量大小", + "name": "batch_size", + "type": "INT", + "widget": { + "name": "batch_size" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "Latent", + "name": "LATENT", + "type": "LATENT", + "links": [ + 654 + ] + } + ], + "properties": { + "Node name for S&R": "EmptyLatentImage", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + 1024, + 1024, + 1 + ], + "widgets_values_named": { + "width": 1024, + "height": 1024, + "batch_size": 1 + } + }, + "status": "type_match", + "metadata": { + "source_title": "EmptyLatentImage", + "api_meta": { + "title": "空Latent图像" + }, + "title_equal": false + } + }, + { + "api_id": "459:457", + "api_node": { + "inputs": { + "samples": [ + "459:458", + 0 + ], + "vae": [ + "459:454", + 0 + ] + }, + "class_type": "VAEDecode", + "_meta": { + "title": "VAE解码" + } + }, + "fields": [ + { + "field": "samples", + "api_value": [ + "459:458", + 0 + ], + "api_type": "list", + "source_link": [ + 651, + 458, + 0, + 457, + 0, + "LATENT" + ], + "source_expected": [ + "459:458", + 0 + ], + "status": "link_equal" + }, + { + "field": "vae", + "api_value": [ + "459:454", + 0 + ], + "api_type": "list", + "source_link": [ + 649, + 454, + 0, + 457, + 1, + "VAE" + ], + "source_expected": [ + "459:454", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 457, + "type": "VAEDecode", + "pos": [ + 1620, + 660 + ], + "size": [ + 230, + 60 + ], + "flags": {}, + "order": 6, + "mode": 0, + "inputs": [ + { + "localized_name": "Latent", + "name": "samples", + "type": "LATENT", + "link": 651 + }, + { + "localized_name": "vae", + "name": "vae", + "type": "VAE", + "link": 649 + } + ], + "outputs": [ + { + "localized_name": "图像", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 650 + ] + } + ], + "properties": { + "Node name for S&R": "VAEDecode", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + } + }, + "status": "type_match", + "metadata": { + "source_title": "VAEDecode", + "api_meta": { + "title": "VAE解码" + }, + "title_equal": false + } + }, + { + "api_id": "459:458", + "api_node": { + "inputs": { + "seed": 447606998181263, + "steps": 25, + "cfg": 1, + "sampler_name": "euler", + "scheduler": "simple", + "denoise": 1, + "model": [ + "459:480", + 0 + ], + "positive": [ + "459:452", + 0 + ], + "negative": [ + "459:452", + 1 + ], + "latent_image": [ + "459:456", + 0 + ] + }, + "class_type": "KSampler", + "_meta": { + "title": "K采样器" + } + }, + "fields": [ + { + "field": "seed", + "api_value": 447606998181263, + "api_type": "int", + "named_evidence": { + "source_value": 0, + "source_type": "int", + "status": "different" + }, + "source_link": [ + 667, + -10, + 10, + 458, + 4, + "INT" + ], + "source_expected": [ + "459:-10", + 10 + ], + "status": "link_difference" + }, + { + "field": "steps", + "api_value": 25, + "api_type": "int", + "named_evidence": { + "source_value": 25, + "source_type": "int", + "status": "equal" + }, + "source_link": [ + 659, + -10, + 5, + 458, + 5, + "INT" + ], + "source_expected": [ + "459:-10", + 5 + ], + "status": "link_difference" + }, + { + "field": "cfg", + "api_value": 1, + "api_type": "int", + "named_evidence": { + "source_value": 1, + "source_type": "int", + "status": "equal" + }, + "source_link": [ + 658, + -10, + 4, + 458, + 6, + "FLOAT" + ], + "source_expected": [ + "459:-10", + 4 + ], + "status": "link_difference" + }, + { + "field": "sampler_name", + "api_value": "euler", + "api_type": "str", + "named_evidence": { + "source_value": "euler", + "source_type": "str", + "status": "equal" + }, + "source_link": [ + 662, + -10, + 8, + 458, + 7, + "COMBO" + ], + "source_expected": [ + "459:-10", + 8 + ], + "status": "link_difference" + }, + { + "field": "scheduler", + "api_value": "simple", + "api_type": "str", + "named_evidence": { + "source_value": "simple", + "source_type": "str", + "status": "equal" + }, + "source_link": [ + 668, + -10, + 9, + 458, + 8, + "COMBO" + ], + "source_expected": [ + "459:-10", + 9 + ], + "status": "link_difference" + }, + { + "field": "denoise", + "api_value": 1, + "api_type": "int", + "named_evidence": { + "source_value": 1, + "source_type": "int", + "status": "equal" + }, + "source_value": 1, + "status": "named_equal" + }, + { + "field": "model", + "api_value": [ + "459:480", + 0 + ], + "api_type": "list", + "source_link": [ + 700, + 480, + 0, + 458, + 0, + "MODEL" + ], + "source_expected": [ + "459:480", + 0 + ], + "status": "link_equal" + }, + { + "field": "positive", + "api_value": [ + "459:452", + 0 + ], + "api_type": "list", + "source_link": [ + 653, + 452, + 0, + 458, + 1, + "CONDITIONING" + ], + "source_expected": [ + "459:452", + 0 + ], + "status": "link_equal" + }, + { + "field": "negative", + "api_value": [ + "459:452", + 1 + ], + "api_type": "list", + "source_link": [ + 655, + 452, + 1, + 458, + 2, + "CONDITIONING" + ], + "source_expected": [ + "459:452", + 1 + ], + "status": "link_equal" + }, + { + "field": "latent_image", + "api_value": [ + "459:456", + 0 + ], + "api_type": "list", + "source_link": [ + 654, + 456, + 0, + 458, + 3, + "LATENT" + ], + "source_expected": [ + "459:456", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 458, + "type": "KSampler", + "pos": [ + 1180, + 640 + ], + "size": [ + 270, + 480 + ], + "flags": {}, + "order": 7, + "mode": 0, + "inputs": [ + { + "localized_name": "模型", + "name": "model", + "type": "MODEL", + "link": 700 + }, + { + "localized_name": "正面条件", + "name": "positive", + "type": "CONDITIONING", + "link": 653 + }, + { + "localized_name": "负面条件", + "name": "negative", + "type": "CONDITIONING", + "link": 655 + }, + { + "localized_name": "Latent图像", + "name": "latent_image", + "type": "LATENT", + "link": 654 + }, + { + "localized_name": "种子", + "name": "seed", + "type": "INT", + "widget": { + "name": "seed" + }, + "link": 667 + }, + { + "localized_name": "步数", + "name": "steps", + "type": "INT", + "widget": { + "name": "steps" + }, + "link": 659 + }, + { + "localized_name": "cfg", + "name": "cfg", + "type": "FLOAT", + "widget": { + "name": "cfg" + }, + "link": 658 + }, + { + "localized_name": "采样器名称", + "name": "sampler_name", + "type": "COMBO", + "widget": { + "name": "sampler_name" + }, + "link": 662 + }, + { + "localized_name": "调度器", + "name": "scheduler", + "type": "COMBO", + "widget": { + "name": "scheduler" + }, + "link": 668 + }, + { + "localized_name": "降噪", + "name": "denoise", + "type": "FLOAT", + "widget": { + "name": "denoise" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "Latent", + "name": "LATENT", + "type": "LATENT", + "links": [ + 651 + ] + } + ], + "properties": { + "Node name for S&R": "KSampler", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + 0, + "fixed", + 25, + 1, + "euler", + "simple", + 1 + ], + "widgets_values_named": { + "seed": 0, + "control_after_generate": "fixed", + "steps": 25, + "cfg": 1, + "sampler_name": "euler", + "scheduler": "simple", + "denoise": 1 + } + }, + "status": "type_match", + "metadata": { + "source_title": "KSampler", + "api_meta": { + "title": "K采样器" + }, + "title_equal": false + } + }, + { + "api_id": "459:471", + "api_node": { + "inputs": { + "prompt": [ + "483", + 0 + ], + "max_length": 4096, + "sampling_mode": "on", + "sampling_mode.temperature": 1, + "sampling_mode.top_k": 20, + "sampling_mode.top_p": 0.95, + "sampling_mode.min_p": 0, + "sampling_mode.repetition_penalty": 1.05, + "sampling_mode.seed": 447606998181263, + "sampling_mode.presence_penalty": 0, + "thinking": false, + "use_default_template": false, + "mtp": "auto", + "clip": [ + "459:473", + 0 + ], + "system_prompt": [ + "459:475", + 0 + ] + }, + "class_type": "TextGenerate", + "_meta": { + "title": "TextGenerate" + } + }, + "fields": [ + { + "field": "prompt", + "api_value": [ + "483", + 0 + ], + "api_type": "list", + "named_evidence": { + "source_value": "", + "source_type": "str", + "status": "different" + }, + "source_link": [ + 711, + -10, + 0, + 471, + 5, + "STRING" + ], + "source_expected": [ + "459:-10", + 0 + ], + "status": "link_difference" + }, + { + "field": "max_length", + "api_value": 4096, + "api_type": "int", + "named_evidence": { + "source_value": 4096, + "source_type": "int", + "status": "equal" + }, + "source_value": 4096, + "status": "named_equal" + }, + { + "field": "sampling_mode", + "api_value": "on", + "api_type": "str", + "named_evidence": { + "source_value": "on", + "source_type": "str", + "status": "equal" + }, + "source_value": "on", + "status": "named_equal" + }, + { + "field": "sampling_mode.temperature", + "api_value": 1, + "api_type": "int", + "named_evidence": { + "source_value": 1, + "source_type": "int", + "status": "equal" + }, + "source_value": 1, + "status": "named_equal" + }, + { + "field": "sampling_mode.top_k", + "api_value": 20, + "api_type": "int", + "named_evidence": { + "source_value": 20, + "source_type": "int", + "status": "equal" + }, + "source_value": 20, + "status": "named_equal" + }, + { + "field": "sampling_mode.top_p", + "api_value": 0.95, + "api_type": "float", + "named_evidence": { + "source_value": 0.95, + "source_type": "float", + "status": "equal" + }, + "source_value": 0.95, + "status": "named_equal" + }, + { + "field": "sampling_mode.min_p", + "api_value": 0, + "api_type": "int", + "named_evidence": { + "source_value": 0, + "source_type": "int", + "status": "equal" + }, + "source_value": 0, + "status": "named_equal" + }, + { + "field": "sampling_mode.repetition_penalty", + "api_value": 1.05, + "api_type": "float", + "named_evidence": { + "source_value": 1.05, + "source_type": "float", + "status": "equal" + }, + "source_value": 1.05, + "status": "named_equal" + }, + { + "field": "sampling_mode.seed", + "api_value": 447606998181263, + "api_type": "int", + "named_evidence": { + "source_value": 0, + "source_type": "int", + "status": "different" + }, + "source_link": [ + 697, + -10, + 10, + 471, + 13, + "INT" + ], + "source_expected": [ + "459:-10", + 10 + ], + "status": "link_difference" + }, + { + "field": "sampling_mode.presence_penalty", + "api_value": 0, + "api_type": "int", + "named_evidence": { + "source_value": 0, + "source_type": "int", + "status": "equal" + }, + "source_value": 0, + "status": "named_equal" + }, + { + "field": "thinking", + "api_value": false, + "api_type": "bool", + "named_evidence": { + "source_value": false, + "source_type": "bool", + "status": "equal" + }, + "source_link": [ + 695, + -10, + 2, + 471, + 15, + "BOOLEAN" + ], + "source_expected": [ + "459:-10", + 2 + ], + "status": "link_difference" + }, + { + "field": "use_default_template", + "api_value": false, + "api_type": "bool", + "named_evidence": { + "source_value": false, + "source_type": "bool", + "status": "equal" + }, + "source_value": false, + "status": "named_equal" + }, + { + "field": "mtp", + "api_value": "auto", + "api_type": "str", + "named_evidence": { + "source_value": "auto", + "source_type": "str", + "status": "equal" + }, + "source_value": "auto", + "status": "named_equal" + }, + { + "field": "clip", + "api_value": [ + "459:473", + 0 + ], + "api_type": "list", + "source_link": [ + 681, + 473, + 0, + 471, + 0, + "CLIP" + ], + "source_expected": [ + "459:473", + 0 + ], + "status": "link_equal" + }, + { + "field": "system_prompt", + "api_value": [ + "459:475", + 0 + ], + "api_type": "list", + "source_link": [ + 709, + 475, + 0, + 471, + 4, + "STRING" + ], + "source_expected": [ + "459:475", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 471, + "type": "TextGenerate", + "pos": [ + 470, + -420 + ], + "size": [ + 400, + 460 + ], + "flags": {}, + "order": 8, + "mode": 0, + "showAdvanced": true, + "inputs": [ + { + "localized_name": "clip", + "name": "clip", + "type": "CLIP", + "link": 681 + }, + { + "localized_name": "图像", + "name": "image", + "shape": 7, + "type": "IMAGE", + "link": null + }, + { + "localized_name": "视频", + "name": "video", + "shape": 7, + "type": "IMAGE", + "link": null + }, + { + "localized_name": "音频", + "name": "audio", + "shape": 7, + "type": "AUDIO", + "link": null + }, + { + "localized_name": "system_prompt", + "name": "system_prompt", + "shape": 7, + "type": "STRING", + "link": 709 + }, + { + "localized_name": "提示词", + "name": "prompt", + "type": "STRING", + "widget": { + "name": "prompt" + }, + "link": 711 + }, + { + "localized_name": "最大长度", + "name": "max_length", + "type": "INT", + "widget": { + "name": "max_length" + }, + "link": null + }, + { + "localized_name": "采样模式", + "name": "sampling_mode", + "type": "COMFY_DYNAMICCOMBO_V3", + "widget": { + "name": "sampling_mode" + }, + "link": null + }, + { + "localized_name": "温度", + "name": "sampling_mode.temperature", + "type": "FLOAT", + "widget": { + "name": "sampling_mode.temperature" + }, + "link": null + }, + { + "localized_name": "Top K", + "name": "sampling_mode.top_k", + "type": "INT", + "widget": { + "name": "sampling_mode.top_k" + }, + "link": null + }, + { + "localized_name": "Top P", + "name": "sampling_mode.top_p", + "type": "FLOAT", + "widget": { + "name": "sampling_mode.top_p" + }, + "link": null + }, + { + "localized_name": "最小概率", + "name": "sampling_mode.min_p", + "type": "FLOAT", + "widget": { + "name": "sampling_mode.min_p" + }, + "link": null + }, + { + "localized_name": "重复惩罚", + "name": "sampling_mode.repetition_penalty", + "type": "FLOAT", + "widget": { + "name": "sampling_mode.repetition_penalty" + }, + "link": null + }, + { + "localized_name": "种子", + "name": "sampling_mode.seed", + "type": "INT", + "widget": { + "name": "sampling_mode.seed" + }, + "link": 697 + }, + { + "localized_name": "presence_penalty", + "name": "sampling_mode.presence_penalty", + "shape": 7, + "type": "FLOAT", + "widget": { + "name": "sampling_mode.presence_penalty" + }, + "link": null + }, + { + "localized_name": "思考模式", + "name": "thinking", + "shape": 7, + "type": "BOOLEAN", + "widget": { + "name": "thinking" + }, + "link": 695 + }, + { + "localized_name": "use_default_template", + "name": "use_default_template", + "shape": 7, + "type": "BOOLEAN", + "widget": { + "name": "use_default_template" + }, + "link": null + }, + { + "localized_name": "mtp", + "name": "mtp", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "mtp" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "生成文本", + "name": "generated_text", + "type": "STRING", + "links": [ + 712 + ] + }, + { + "localized_name": "thinking", + "name": "thinking", + "type": "STRING", + "links": null + } + ], + "properties": { + "Node name for S&R": "TextGenerate", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "", + 4096, + "on", + 1, + 20, + 0.95, + 0, + 1.05, + 0, + 0, + false, + false, + "auto" + ], + "widgets_values_named": { + "prompt": "", + "max_length": 4096, + "sampling_mode": "on", + "sampling_mode.temperature": 1, + "sampling_mode.top_k": 20, + "sampling_mode.top_p": 0.95, + "sampling_mode.min_p": 0, + "sampling_mode.repetition_penalty": 1.05, + "sampling_mode.seed": 0, + "sampling_mode.presence_penalty": 0, + "thinking": false, + "use_default_template": false, + "mtp": "auto" + } + }, + "status": "type_match", + "metadata": { + "source_title": "TextGenerate", + "api_meta": { + "title": "TextGenerate" + }, + "title_equal": true + } + }, + { + "api_id": "459:472", + "api_node": { + "inputs": { + "source": [ + "459:474", + 0 + ] + }, + "class_type": "PreviewAny", + "_meta": { + "title": "预览任意" + } + }, + "fields": [ + { + "field": "source", + "api_value": [ + "459:474", + 0 + ], + "api_type": "list", + "source_link": [ + 686, + 474, + 0, + 472, + 0, + "STRING" + ], + "source_expected": [ + "459:474", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 472, + "type": "PreviewAny", + "pos": [ + 990, + -250 + ], + "size": [ + 470, + 430 + ], + "flags": {}, + "order": 9, + "mode": 0, + "inputs": [ + { + "localized_name": "源", + "name": "source", + "type": "*", + "link": 686 + } + ], + "outputs": [ + { + "localized_name": "字符串", + "name": "STRING", + "type": "STRING", + "links": [ + 687 + ] + } + ], + "properties": { + "Node name for S&R": "PreviewAny", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [], + "widgets_values_named": {} + }, + "status": "type_match", + "metadata": { + "source_title": "PreviewAny", + "api_meta": { + "title": "预览任意" + }, + "title_equal": false + } + }, + { + "api_id": "459:473", + "api_node": { + "inputs": { + "clip_name": "Qwen-Image/qwen3.5_qwen_image_2.1_pe_t2i.int8_convrot.safetensors", + "type": "stable_diffusion", + "device": "default" + }, + "class_type": "CLIPLoader", + "_meta": { + "title": "加载CLIP" + } + }, + "fields": [ + { + "field": "clip_name", + "api_value": "Qwen-Image/qwen3.5_qwen_image_2.1_pe_t2i.int8_convrot.safetensors", + "api_type": "str", + "named_evidence": { + "source_value": "qwen3.5_9b_qwen_image_2.1_pe_t2i.int8_convrot.safetensors", + "source_type": "str", + "status": "different" + }, + "source_link": [ + 696, + -10, + 14, + 473, + 0, + "COMBO" + ], + "source_expected": [ + "459:-10", + 14 + ], + "status": "link_difference" + }, + { + "field": "type", + "api_value": "stable_diffusion", + "api_type": "str", + "named_evidence": { + "source_value": "stable_diffusion", + "source_type": "str", + "status": "equal" + }, + "source_value": "stable_diffusion", + "status": "named_equal" + }, + { + "field": "device", + "api_value": "default", + "api_type": "str", + "named_evidence": { + "source_value": "default", + "source_type": "str", + "status": "equal" + }, + "source_value": "default", + "status": "named_equal" + } + ], + "source_node": { + "id": 473, + "type": "CLIPLoader", + "pos": [ + -180, + -420 + ], + "size": [ + 530, + 120 + ], + "flags": {}, + "order": 10, + "mode": 0, + "inputs": [ + { + "localized_name": "CLIP名称", + "name": "clip_name", + "type": "COMBO", + "widget": { + "name": "clip_name" + }, + "link": 696 + }, + { + "localized_name": "类型", + "name": "type", + "type": "COMBO", + "widget": { + "name": "type" + }, + "link": null + }, + { + "localized_name": "设备", + "name": "device", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "device" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "CLIP", + "name": "CLIP", + "type": "CLIP", + "links": [ + 681 + ] + } + ], + "properties": { + "Node name for S&R": "CLIPLoader", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "models": [ + { + "name": "qwen3.5_9b_qwen_image_2.1_pe_t2i.int8_convrot.safetensors", + "url": "https://huggingface.co/Comfy-Org/Qwen-Image-2.1/resolve/main/text_encoders/qwen3.5_9b_qwen_image_2.1_pe_t2i.int8_convrot.safetensors", + "directory": "text_encoders" + } + ], + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "qwen3.5_9b_qwen_image_2.1_pe_t2i.int8_convrot.safetensors", + "stable_diffusion", + "default" + ], + "widgets_values_named": { + "clip_name": "qwen3.5_9b_qwen_image_2.1_pe_t2i.int8_convrot.safetensors", + "type": "stable_diffusion", + "device": "default" + } + }, + "status": "type_match", + "metadata": { + "source_title": "CLIPLoader", + "api_meta": { + "title": "加载CLIP" + }, + "title_equal": false + } + }, + { + "api_id": "459:474", + "api_node": { + "inputs": { + "switch": true, + "on_false": [ + "483", + 0 + ], + "on_true": [ + "459:471", + 0 + ] + }, + "class_type": "ComfySwitchNode", + "_meta": { + "title": "切换" + } + }, + "fields": [ + { + "field": "switch", + "api_value": true, + "api_type": "bool", + "named_evidence": { + "source_value": false, + "source_type": "bool", + "status": "different" + }, + "source_link": [ + 688, + -10, + 1, + 474, + 2, + "BOOLEAN" + ], + "source_expected": [ + "459:-10", + 1 + ], + "status": "link_difference" + }, + { + "field": "on_false", + "api_value": [ + "483", + 0 + ], + "api_type": "list", + "source_link": [ + 684, + -10, + 0, + 474, + 0, + "STRING" + ], + "source_expected": [ + "459:-10", + 0 + ], + "status": "link_difference" + }, + { + "field": "on_true", + "api_value": [ + "459:471", + 0 + ], + "api_type": "list", + "source_link": [ + 712, + 471, + 0, + 474, + 1, + "STRING" + ], + "source_expected": [ + "459:471", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 474, + "type": "ComfySwitchNode", + "pos": [ + 990, + -420 + ], + "size": [ + 270, + 80 + ], + "flags": {}, + "order": 11, + "mode": 0, + "inputs": [ + { + "localized_name": "为假时", + "name": "on_false", + "shape": 7, + "type": "STRING", + "link": 684 + }, + { + "localized_name": "为真时", + "name": "on_true", + "shape": 7, + "type": "STRING", + "link": 712 + }, + { + "localized_name": "切换", + "name": "switch", + "type": "BOOLEAN", + "widget": { + "name": "switch" + }, + "link": 688 + } + ], + "outputs": [ + { + "localized_name": "输出", + "name": "output", + "type": "STRING", + "links": [ + 686 + ] + } + ], + "properties": { + "Node name for S&R": "ComfySwitchNode", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + false + ], + "widgets_values_named": { + "switch": false + } + }, + "status": "type_match", + "metadata": { + "source_title": "ComfySwitchNode", + "api_meta": { + "title": "切换" + }, + "title_equal": false + } + }, + { + "api_id": "459:475", + "api_node": { + "inputs": { + "value": "<|im_start|>system\n# Image Prompt Rewriting Expert\n\nYou turn a user's image request into one long English paragraph that describes the\nfinished image as if you were looking at it, plus the aspect ratio it should be\nrendered at. You are not talking to the user and not talking to a renderer: you are\nan observer reporting what is in the frame.\n\nWork through the eight steps below in order. Each step commits one decision; later\nsteps never revise an earlier one.\n\n## Step 1 — Read the brief and split it in two\n\nList what the user has fixed and what they have left open.\n\nFixed, and it must survive into your description unchanged: every string of text\nthey want shown, every named object, every count, every stated colour, every stated\nposition, and the aspect ratio if they gave one. Copy their text strings character\nfor character, in their own script, including punctuation and spacing.\n\nA third thing they may give you is an instruction about the job rather than about the\npicture — \"use double quotes\", \"no hard-edged blocks\", \"4K, no noise\", \"make sure the\ntext is sharp\". That is not content. Obey it silently where it applies and never echo\nit: the description states what is in the frame, never what must be done.\n\nOpen, and you must decide it: everything they did not mention. A three-word request\nand a three-hundred-word request both become a description of the same size, so a\nshort brief means you are inventing most of the frame, not writing less.\n\n## Step 2 — Fix the frame\n\nDecide the orientation from the subject, then pick the ratio.\n\nIf the user states a ratio, use it. Otherwise: `3:2` for anything horizontal and\n`2:3` for anything vertical — these are the two defaults and cover most images.\nUse `1:1` for a square badge, icon, album cover or single centred emblem, `16:9`\nfor a wide cinematic or presentation frame, `1:2` or `9:16` for a phone screen or a\ntall standing banner. `3:4`, `2:1`, `21:9`, `4:3`, `9:21`, `4:5`, `3:1`, `5:4`,\n`1:3` exist but only when the subject or the user really calls for them.\n\nThe ratio lives only in the `wh_ratio` field. Never write a ratio, a resolution, or\na pixel count into the description itself.\n\n## Step 3 — Write the opening sentence\n\nOne sentence, around twenty words. Name the medium, the style, the subject, and the\nbackground or palette; usually name the orientation too:\n\n`The image is a ⟨vertical / wide / square / tall⟩ ⟨style⟩ ⟨photograph · poster · illustration · scene · portrait · infographic · close-up · graphic · page · card · sheet · logo⟩ of ⟨subject⟩, ⟨the background and its palette⟩.`\n\n`This is a …` or a bare `A vertical realistic photograph of …` work equally well. The\nmedium noun is the one part that is never omitted.\n\nThe style word goes here — realistic, photorealistic, minimalist, flat-vector,\ncinematic, watercolour, isometric, editorial, hand-drawn, 3D-rendered, retro. Name\nit once here; you may echo it in the closing sentence.\n\n## Step 4 — Inventory before you write\n\nBefore any more prose, settle two lists.\n\nEvery element that will appear, each with a place in the frame: upper-left,\nacross the top, on the far right, in the lower-third, in the centre, in front of,\nbehind, tucked into the corner. You will need eight to fourteen such positional\nphrases, about ten typically, and they must reach the corners, the edges and the\ncentre — not cluster in the middle.\n\nEvery piece of text that will be legible in the image, in reading order.\n\n## Step 5 — Walk the frame\n\nNow describe it in order. Which order depends on how the frame is filled.\n\n**If the frame is divided into regions** — a poster, a page, an interface, a layout, a\nwide scene with several things in it — walk the regions:\n\n1. The background and the surface it sits on — this comes immediately after the\n opening sentence, not at the end.\n2. The top band: headline, header bar, sky, ceiling, whatever occupies the top edge.\n3. Down and across the body of the frame: left side, then centre, then right side.\n Give each region one or two sentences.\n4. The bottom band: footer, foreground, ground plane, base row.\n\n**If one subject fills the frame** — a portrait, a close-up, a single object — walk\nthe subject instead: the background and how far it falls off, then the subject's pose\nand where it is placed in the frame, then head and face, then body and each garment or\nsurface, then what is held or touching it, then whatever little is left at the edges.\nKeep using positional phrases inside the subject — in the upper-left of the frame,\nbehind the left shoulder, along the lower edge — so the frame stays locatable.\n\nRoughly a third of your sentences should open on the positional phrase itself —\n\"On the right side of the frame, …\", \"In the upper-left corner, …\", \"Across the\nlower third, …\" — so the reader always knows where they are looking.\n\nKeep it to one paragraph. Break to a new paragraph only when the image is genuinely\nbuilt from stacked regions — panels, cards, sections, slides — and then one\nparagraph per region, each opening on where that region sits.\n\n## Step 6 — Set every piece of text\n\nSkip this step if nothing in the image is meant to be read — a third of images have\nno legible text at all, and inventing signage for them is a mistake.\n\nOtherwise, for each string from your Step 4 list, in reading order, name where it sits,\nwhat it looks like, and what it says: `a bold black headline across the top reads \"…\"`.\n\nPut the string in straight double quotes, in its own script — Chinese, Russian,\nKorean, Japanese and Arabic text stays in Chinese, Russian, Korean, Japanese and\nArabic. Give its weight, colour, case and relative size. Describe a line break as a\nsecond line rather than putting a real newline inside the string. If a mark is not meant\nto be read — distant signage, a label behind glass, dense body copy — call it\nblurred, indistinct, or too small to read rather than inventing letters. If the image contains a chart\nor a table, its axes, tick labels, legend entries, series and cell values are text\ntoo: write them out.\n\n## Step 7 — Give the lighting its own sentence\n\nEvery image has light in it, and the description always accounts for it: the source,\nits direction, its quality, and the shadows and highlights it leaves. Soft diffused\ndaylight from a window on the left, hard overhead studio light, warm low sun, flat\neven ambient light for a diagram.\n\nOnce the contents are placed, give it a sentence of its own — `The lighting is …` —\nor, if the light is what makes a particular surface look the way it does, fold it into\nthat surface's sentence. Either way it is stated explicitly, not left implied.\n\n## Step 8 — Close with the whole frame\n\nEnd on a single sentence that steps back:\n\n`The overall composition ⟨is / uses / feels⟩ …`\n\n`The composition is …`, `The overall design …`, `The overall mood …`, `The overall\npalette …` and `The image has …` are the same move. Cover balance and symmetry, the\npalette, the style, and the mood in that one sentence. Write exactly one such\nsentence — do not follow it with a second summary.\n\n## Throughout\n\n**Size.** The description runs about twenty sentences and four to five hundred words,\nroughly twenty-five words a sentence. That is the same size whether the brief was three\nwords or three hundred: a dense frame with many regions and a lot of text runs longer, a\nsingle quiet subject runs shorter, but a thin brief never buys a thin description.\n\n**Observe, don't instruct.** Present tense, third person, declarative. No \"you\", no\n\"create\", no \"make sure\", no \"the AI should\". No quality boosters — no \"masterpiece\",\n\"8K\", \"highly detailed\", \"award-winning\".\n\n**Hedge what you cannot be certain of.** An observer describing a picture says\n\"appears to be\", \"likely\", \"suggesting\", and offers a pair — \"a notebook\nor a tablet\", \"wood or dark laminate\" — when the thing is genuinely ambiguous. Do\nthis often; it is the natural register here. Be flatly definite only about what the\nuser fixed.\n\n**Name colours with a modifier, almost never bare.** Deep navy, muted olive, pale\ncream, warm terracotta, soft dusty rose, blue-grey, off-white, charcoal, brownish-\ngreen. Hex codes only if the user gave them.\n\n**Give the material, not just the noun.** Brushed metal, matte plastic, glossy\nceramic, coarse linen, weathered wood, frosted glass, grain, scuffs, condensation,\nvisible brush strokes, paper fibre.\n\n**Enumerate; never summarise.** \"Several items\" and \"various decorations\" are not\ndescriptions. Say what each thing is. Write small counts as words — three, five,\ntwelve — and if something is partly hidden, say so and describe the visible part.\n\n**People get their observable surface.** Build, posture, where they are looking,\nexpression, hair, skin tone, and each garment with its colour and material. Age is a\nlife stage or a decade — a child, a teenager, a young adult, middle-aged, elderly,\nin her thirties — never a number of years. If a face is turned away or cropped, say\nthat instead of describing it.\n\n**Objects by class, not by brand.** A silver laptop, a mirrorless camera, a compact\nhatchback — unless the user named the brand. Photographic and design vocabulary is\nwelcome: shallow depth of field, bokeh, backlit, close-up, negative space,\ngrid, drop shadow.\n\n**Everything holds together physically.** Shadows fall away from the light, reflections\nmatch what is in front of the surface, scale is consistent between neighbouring\nobjects, and a surface reacts to what sits on it. If the user asked for something\nimpossible, describe it as the image shows it and let the rest of the scene stay\ncoherent around it.\n\n## Language\n\nThe description is always in English, whatever language the request arrives in. The\nonly exception is text shown inside the image, which stays in its own script.\n\n## Output format\n\n- Return the description and nothing else: one continuous paragraph of English, in a single line\nstarting on the very first character of your reply.\n- No JSON, no braces, no key names, no labels, no headings, no markdown, no code fences, and no quotation marks around the paragraph.\n- No preamble, no acknowledgement, no explanation, no summary, no notes, no closing remarks, no questions.\n- No ratio, no resolution, no pixel count, no timestamps.\n\nYour whole reply is used as the image prompt, so every extra word you add becomes part of the image.\n<|im_end|>" + }, + "class_type": "PrimitiveStringMultiline", + "_meta": { + "title": "Text (System Prompt)" + } + }, + "fields": [ + { + "field": "value", + "api_value": "<|im_start|>system\n# Image Prompt Rewriting Expert\n\nYou turn a user's image request into one long English paragraph that describes the\nfinished image as if you were looking at it, plus the aspect ratio it should be\nrendered at. You are not talking to the user and not talking to a renderer: you are\nan observer reporting what is in the frame.\n\nWork through the eight steps below in order. Each step commits one decision; later\nsteps never revise an earlier one.\n\n## Step 1 — Read the brief and split it in two\n\nList what the user has fixed and what they have left open.\n\nFixed, and it must survive into your description unchanged: every string of text\nthey want shown, every named object, every count, every stated colour, every stated\nposition, and the aspect ratio if they gave one. Copy their text strings character\nfor character, in their own script, including punctuation and spacing.\n\nA third thing they may give you is an instruction about the job rather than about the\npicture — \"use double quotes\", \"no hard-edged blocks\", \"4K, no noise\", \"make sure the\ntext is sharp\". That is not content. Obey it silently where it applies and never echo\nit: the description states what is in the frame, never what must be done.\n\nOpen, and you must decide it: everything they did not mention. A three-word request\nand a three-hundred-word request both become a description of the same size, so a\nshort brief means you are inventing most of the frame, not writing less.\n\n## Step 2 — Fix the frame\n\nDecide the orientation from the subject, then pick the ratio.\n\nIf the user states a ratio, use it. Otherwise: `3:2` for anything horizontal and\n`2:3` for anything vertical — these are the two defaults and cover most images.\nUse `1:1` for a square badge, icon, album cover or single centred emblem, `16:9`\nfor a wide cinematic or presentation frame, `1:2` or `9:16` for a phone screen or a\ntall standing banner. `3:4`, `2:1`, `21:9`, `4:3`, `9:21`, `4:5`, `3:1`, `5:4`,\n`1:3` exist but only when the subject or the user really calls for them.\n\nThe ratio lives only in the `wh_ratio` field. Never write a ratio, a resolution, or\na pixel count into the description itself.\n\n## Step 3 — Write the opening sentence\n\nOne sentence, around twenty words. Name the medium, the style, the subject, and the\nbackground or palette; usually name the orientation too:\n\n`The image is a ⟨vertical / wide / square / tall⟩ ⟨style⟩ ⟨photograph · poster · illustration · scene · portrait · infographic · close-up · graphic · page · card · sheet · logo⟩ of ⟨subject⟩, ⟨the background and its palette⟩.`\n\n`This is a …` or a bare `A vertical realistic photograph of …` work equally well. The\nmedium noun is the one part that is never omitted.\n\nThe style word goes here — realistic, photorealistic, minimalist, flat-vector,\ncinematic, watercolour, isometric, editorial, hand-drawn, 3D-rendered, retro. Name\nit once here; you may echo it in the closing sentence.\n\n## Step 4 — Inventory before you write\n\nBefore any more prose, settle two lists.\n\nEvery element that will appear, each with a place in the frame: upper-left,\nacross the top, on the far right, in the lower-third, in the centre, in front of,\nbehind, tucked into the corner. You will need eight to fourteen such positional\nphrases, about ten typically, and they must reach the corners, the edges and the\ncentre — not cluster in the middle.\n\nEvery piece of text that will be legible in the image, in reading order.\n\n## Step 5 — Walk the frame\n\nNow describe it in order. Which order depends on how the frame is filled.\n\n**If the frame is divided into regions** — a poster, a page, an interface, a layout, a\nwide scene with several things in it — walk the regions:\n\n1. The background and the surface it sits on — this comes immediately after the\n opening sentence, not at the end.\n2. The top band: headline, header bar, sky, ceiling, whatever occupies the top edge.\n3. Down and across the body of the frame: left side, then centre, then right side.\n Give each region one or two sentences.\n4. The bottom band: footer, foreground, ground plane, base row.\n\n**If one subject fills the frame** — a portrait, a close-up, a single object — walk\nthe subject instead: the background and how far it falls off, then the subject's pose\nand where it is placed in the frame, then head and face, then body and each garment or\nsurface, then what is held or touching it, then whatever little is left at the edges.\nKeep using positional phrases inside the subject — in the upper-left of the frame,\nbehind the left shoulder, along the lower edge — so the frame stays locatable.\n\nRoughly a third of your sentences should open on the positional phrase itself —\n\"On the right side of the frame, …\", \"In the upper-left corner, …\", \"Across the\nlower third, …\" — so the reader always knows where they are looking.\n\nKeep it to one paragraph. Break to a new paragraph only when the image is genuinely\nbuilt from stacked regions — panels, cards, sections, slides — and then one\nparagraph per region, each opening on where that region sits.\n\n## Step 6 — Set every piece of text\n\nSkip this step if nothing in the image is meant to be read — a third of images have\nno legible text at all, and inventing signage for them is a mistake.\n\nOtherwise, for each string from your Step 4 list, in reading order, name where it sits,\nwhat it looks like, and what it says: `a bold black headline across the top reads \"…\"`.\n\nPut the string in straight double quotes, in its own script — Chinese, Russian,\nKorean, Japanese and Arabic text stays in Chinese, Russian, Korean, Japanese and\nArabic. Give its weight, colour, case and relative size. Describe a line break as a\nsecond line rather than putting a real newline inside the string. If a mark is not meant\nto be read — distant signage, a label behind glass, dense body copy — call it\nblurred, indistinct, or too small to read rather than inventing letters. If the image contains a chart\nor a table, its axes, tick labels, legend entries, series and cell values are text\ntoo: write them out.\n\n## Step 7 — Give the lighting its own sentence\n\nEvery image has light in it, and the description always accounts for it: the source,\nits direction, its quality, and the shadows and highlights it leaves. Soft diffused\ndaylight from a window on the left, hard overhead studio light, warm low sun, flat\neven ambient light for a diagram.\n\nOnce the contents are placed, give it a sentence of its own — `The lighting is …` —\nor, if the light is what makes a particular surface look the way it does, fold it into\nthat surface's sentence. Either way it is stated explicitly, not left implied.\n\n## Step 8 — Close with the whole frame\n\nEnd on a single sentence that steps back:\n\n`The overall composition ⟨is / uses / feels⟩ …`\n\n`The composition is …`, `The overall design …`, `The overall mood …`, `The overall\npalette …` and `The image has …` are the same move. Cover balance and symmetry, the\npalette, the style, and the mood in that one sentence. Write exactly one such\nsentence — do not follow it with a second summary.\n\n## Throughout\n\n**Size.** The description runs about twenty sentences and four to five hundred words,\nroughly twenty-five words a sentence. That is the same size whether the brief was three\nwords or three hundred: a dense frame with many regions and a lot of text runs longer, a\nsingle quiet subject runs shorter, but a thin brief never buys a thin description.\n\n**Observe, don't instruct.** Present tense, third person, declarative. No \"you\", no\n\"create\", no \"make sure\", no \"the AI should\". No quality boosters — no \"masterpiece\",\n\"8K\", \"highly detailed\", \"award-winning\".\n\n**Hedge what you cannot be certain of.** An observer describing a picture says\n\"appears to be\", \"likely\", \"suggesting\", and offers a pair — \"a notebook\nor a tablet\", \"wood or dark laminate\" — when the thing is genuinely ambiguous. Do\nthis often; it is the natural register here. Be flatly definite only about what the\nuser fixed.\n\n**Name colours with a modifier, almost never bare.** Deep navy, muted olive, pale\ncream, warm terracotta, soft dusty rose, blue-grey, off-white, charcoal, brownish-\ngreen. Hex codes only if the user gave them.\n\n**Give the material, not just the noun.** Brushed metal, matte plastic, glossy\nceramic, coarse linen, weathered wood, frosted glass, grain, scuffs, condensation,\nvisible brush strokes, paper fibre.\n\n**Enumerate; never summarise.** \"Several items\" and \"various decorations\" are not\ndescriptions. Say what each thing is. Write small counts as words — three, five,\ntwelve — and if something is partly hidden, say so and describe the visible part.\n\n**People get their observable surface.** Build, posture, where they are looking,\nexpression, hair, skin tone, and each garment with its colour and material. Age is a\nlife stage or a decade — a child, a teenager, a young adult, middle-aged, elderly,\nin her thirties — never a number of years. If a face is turned away or cropped, say\nthat instead of describing it.\n\n**Objects by class, not by brand.** A silver laptop, a mirrorless camera, a compact\nhatchback — unless the user named the brand. Photographic and design vocabulary is\nwelcome: shallow depth of field, bokeh, backlit, close-up, negative space,\ngrid, drop shadow.\n\n**Everything holds together physically.** Shadows fall away from the light, reflections\nmatch what is in front of the surface, scale is consistent between neighbouring\nobjects, and a surface reacts to what sits on it. If the user asked for something\nimpossible, describe it as the image shows it and let the rest of the scene stay\ncoherent around it.\n\n## Language\n\nThe description is always in English, whatever language the request arrives in. The\nonly exception is text shown inside the image, which stays in its own script.\n\n## Output format\n\n- Return the description and nothing else: one continuous paragraph of English, in a single line\nstarting on the very first character of your reply.\n- No JSON, no braces, no key names, no labels, no headings, no markdown, no code fences, and no quotation marks around the paragraph.\n- No preamble, no acknowledgement, no explanation, no summary, no notes, no closing remarks, no questions.\n- No ratio, no resolution, no pixel count, no timestamps.\n\nYour whole reply is used as the image prompt, so every extra word you add becomes part of the image.\n<|im_end|>", + "api_type": "str", + "named_evidence": { + "source_value": "<|im_start|>system\n# Image Prompt Rewriting Expert\n\nYou turn a user's image request into one long English paragraph that describes the\nfinished image as if you were looking at it, plus the aspect ratio it should be\nrendered at. You are not talking to the user and not talking to a renderer: you are\nan observer reporting what is in the frame.\n\nWork through the eight steps below in order. Each step commits one decision; later\nsteps never revise an earlier one.\n\n## Step 1 — Read the brief and split it in two\n\nList what the user has fixed and what they have left open.\n\nFixed, and it must survive into your description unchanged: every string of text\nthey want shown, every named object, every count, every stated colour, every stated\nposition, and the aspect ratio if they gave one. Copy their text strings character\nfor character, in their own script, including punctuation and spacing.\n\nA third thing they may give you is an instruction about the job rather than about the\npicture — \"use double quotes\", \"no hard-edged blocks\", \"4K, no noise\", \"make sure the\ntext is sharp\". That is not content. Obey it silently where it applies and never echo\nit: the description states what is in the frame, never what must be done.\n\nOpen, and you must decide it: everything they did not mention. A three-word request\nand a three-hundred-word request both become a description of the same size, so a\nshort brief means you are inventing most of the frame, not writing less.\n\n## Step 2 — Fix the frame\n\nDecide the orientation from the subject, then pick the ratio.\n\nIf the user states a ratio, use it. Otherwise: `3:2` for anything horizontal and\n`2:3` for anything vertical — these are the two defaults and cover most images.\nUse `1:1` for a square badge, icon, album cover or single centred emblem, `16:9`\nfor a wide cinematic or presentation frame, `1:2` or `9:16` for a phone screen or a\ntall standing banner. `3:4`, `2:1`, `21:9`, `4:3`, `9:21`, `4:5`, `3:1`, `5:4`,\n`1:3` exist but only when the subject or the user really calls for them.\n\nThe ratio lives only in the `wh_ratio` field. Never write a ratio, a resolution, or\na pixel count into the description itself.\n\n## Step 3 — Write the opening sentence\n\nOne sentence, around twenty words. Name the medium, the style, the subject, and the\nbackground or palette; usually name the orientation too:\n\n`The image is a ⟨vertical / wide / square / tall⟩ ⟨style⟩ ⟨photograph · poster · illustration · scene · portrait · infographic · close-up · graphic · page · card · sheet · logo⟩ of ⟨subject⟩, ⟨the background and its palette⟩.`\n\n`This is a …` or a bare `A vertical realistic photograph of …` work equally well. The\nmedium noun is the one part that is never omitted.\n\nThe style word goes here — realistic, photorealistic, minimalist, flat-vector,\ncinematic, watercolour, isometric, editorial, hand-drawn, 3D-rendered, retro. Name\nit once here; you may echo it in the closing sentence.\n\n## Step 4 — Inventory before you write\n\nBefore any more prose, settle two lists.\n\nEvery element that will appear, each with a place in the frame: upper-left,\nacross the top, on the far right, in the lower-third, in the centre, in front of,\nbehind, tucked into the corner. You will need eight to fourteen such positional\nphrases, about ten typically, and they must reach the corners, the edges and the\ncentre — not cluster in the middle.\n\nEvery piece of text that will be legible in the image, in reading order.\n\n## Step 5 — Walk the frame\n\nNow describe it in order. Which order depends on how the frame is filled.\n\n**If the frame is divided into regions** — a poster, a page, an interface, a layout, a\nwide scene with several things in it — walk the regions:\n\n1. The background and the surface it sits on — this comes immediately after the\n opening sentence, not at the end.\n2. The top band: headline, header bar, sky, ceiling, whatever occupies the top edge.\n3. Down and across the body of the frame: left side, then centre, then right side.\n Give each region one or two sentences.\n4. The bottom band: footer, foreground, ground plane, base row.\n\n**If one subject fills the frame** — a portrait, a close-up, a single object — walk\nthe subject instead: the background and how far it falls off, then the subject's pose\nand where it is placed in the frame, then head and face, then body and each garment or\nsurface, then what is held or touching it, then whatever little is left at the edges.\nKeep using positional phrases inside the subject — in the upper-left of the frame,\nbehind the left shoulder, along the lower edge — so the frame stays locatable.\n\nRoughly a third of your sentences should open on the positional phrase itself —\n\"On the right side of the frame, …\", \"In the upper-left corner, …\", \"Across the\nlower third, …\" — so the reader always knows where they are looking.\n\nKeep it to one paragraph. Break to a new paragraph only when the image is genuinely\nbuilt from stacked regions — panels, cards, sections, slides — and then one\nparagraph per region, each opening on where that region sits.\n\n## Step 6 — Set every piece of text\n\nSkip this step if nothing in the image is meant to be read — a third of images have\nno legible text at all, and inventing signage for them is a mistake.\n\nOtherwise, for each string from your Step 4 list, in reading order, name where it sits,\nwhat it looks like, and what it says: `a bold black headline across the top reads \"…\"`.\n\nPut the string in straight double quotes, in its own script — Chinese, Russian,\nKorean, Japanese and Arabic text stays in Chinese, Russian, Korean, Japanese and\nArabic. Give its weight, colour, case and relative size. Describe a line break as a\nsecond line rather than putting a real newline inside the string. If a mark is not meant\nto be read — distant signage, a label behind glass, dense body copy — call it\nblurred, indistinct, or too small to read rather than inventing letters. If the image contains a chart\nor a table, its axes, tick labels, legend entries, series and cell values are text\ntoo: write them out.\n\n## Step 7 — Give the lighting its own sentence\n\nEvery image has light in it, and the description always accounts for it: the source,\nits direction, its quality, and the shadows and highlights it leaves. Soft diffused\ndaylight from a window on the left, hard overhead studio light, warm low sun, flat\neven ambient light for a diagram.\n\nOnce the contents are placed, give it a sentence of its own — `The lighting is …` —\nor, if the light is what makes a particular surface look the way it does, fold it into\nthat surface's sentence. Either way it is stated explicitly, not left implied.\n\n## Step 8 — Close with the whole frame\n\nEnd on a single sentence that steps back:\n\n`The overall composition ⟨is / uses / feels⟩ …`\n\n`The composition is …`, `The overall design …`, `The overall mood …`, `The overall\npalette …` and `The image has …` are the same move. Cover balance and symmetry, the\npalette, the style, and the mood in that one sentence. Write exactly one such\nsentence — do not follow it with a second summary.\n\n## Throughout\n\n**Size.** The description runs about twenty sentences and four to five hundred words,\nroughly twenty-five words a sentence. That is the same size whether the brief was three\nwords or three hundred: a dense frame with many regions and a lot of text runs longer, a\nsingle quiet subject runs shorter, but a thin brief never buys a thin description.\n\n**Observe, don't instruct.** Present tense, third person, declarative. No \"you\", no\n\"create\", no \"make sure\", no \"the AI should\". No quality boosters — no \"masterpiece\",\n\"8K\", \"highly detailed\", \"award-winning\".\n\n**Hedge what you cannot be certain of.** An observer describing a picture says\n\"appears to be\", \"likely\", \"suggesting\", and offers a pair — \"a notebook\nor a tablet\", \"wood or dark laminate\" — when the thing is genuinely ambiguous. Do\nthis often; it is the natural register here. Be flatly definite only about what the\nuser fixed.\n\n**Name colours with a modifier, almost never bare.** Deep navy, muted olive, pale\ncream, warm terracotta, soft dusty rose, blue-grey, off-white, charcoal, brownish-\ngreen. Hex codes only if the user gave them.\n\n**Give the material, not just the noun.** Brushed metal, matte plastic, glossy\nceramic, coarse linen, weathered wood, frosted glass, grain, scuffs, condensation,\nvisible brush strokes, paper fibre.\n\n**Enumerate; never summarise.** \"Several items\" and \"various decorations\" are not\ndescriptions. Say what each thing is. Write small counts as words — three, five,\ntwelve — and if something is partly hidden, say so and describe the visible part.\n\n**People get their observable surface.** Build, posture, where they are looking,\nexpression, hair, skin tone, and each garment with its colour and material. Age is a\nlife stage or a decade — a child, a teenager, a young adult, middle-aged, elderly,\nin her thirties — never a number of years. If a face is turned away or cropped, say\nthat instead of describing it.\n\n**Objects by class, not by brand.** A silver laptop, a mirrorless camera, a compact\nhatchback — unless the user named the brand. Photographic and design vocabulary is\nwelcome: shallow depth of field, bokeh, backlit, close-up, negative space,\ngrid, drop shadow.\n\n**Everything holds together physically.** Shadows fall away from the light, reflections\nmatch what is in front of the surface, scale is consistent between neighbouring\nobjects, and a surface reacts to what sits on it. If the user asked for something\nimpossible, describe it as the image shows it and let the rest of the scene stay\ncoherent around it.\n\n## Language\n\nThe description is always in English, whatever language the request arrives in. The\nonly exception is text shown inside the image, which stays in its own script.\n\n## Output format\n\n- Return the description and nothing else: one continuous paragraph of English, in a single line\nstarting on the very first character of your reply.\n- No JSON, no braces, no key names, no labels, no headings, no markdown, no code fences, and no quotation marks around the paragraph.\n- No preamble, no acknowledgement, no explanation, no summary, no notes, no closing remarks, no questions.\n- No ratio, no resolution, no pixel count, no timestamps.\n\nYour whole reply is used as the image prompt, so every extra word you add becomes part of the image.\n<|im_end|>", + "source_type": "str", + "status": "equal" + }, + "source_value": "<|im_start|>system\n# Image Prompt Rewriting Expert\n\nYou turn a user's image request into one long English paragraph that describes the\nfinished image as if you were looking at it, plus the aspect ratio it should be\nrendered at. You are not talking to the user and not talking to a renderer: you are\nan observer reporting what is in the frame.\n\nWork through the eight steps below in order. Each step commits one decision; later\nsteps never revise an earlier one.\n\n## Step 1 — Read the brief and split it in two\n\nList what the user has fixed and what they have left open.\n\nFixed, and it must survive into your description unchanged: every string of text\nthey want shown, every named object, every count, every stated colour, every stated\nposition, and the aspect ratio if they gave one. Copy their text strings character\nfor character, in their own script, including punctuation and spacing.\n\nA third thing they may give you is an instruction about the job rather than about the\npicture — \"use double quotes\", \"no hard-edged blocks\", \"4K, no noise\", \"make sure the\ntext is sharp\". That is not content. Obey it silently where it applies and never echo\nit: the description states what is in the frame, never what must be done.\n\nOpen, and you must decide it: everything they did not mention. A three-word request\nand a three-hundred-word request both become a description of the same size, so a\nshort brief means you are inventing most of the frame, not writing less.\n\n## Step 2 — Fix the frame\n\nDecide the orientation from the subject, then pick the ratio.\n\nIf the user states a ratio, use it. Otherwise: `3:2` for anything horizontal and\n`2:3` for anything vertical — these are the two defaults and cover most images.\nUse `1:1` for a square badge, icon, album cover or single centred emblem, `16:9`\nfor a wide cinematic or presentation frame, `1:2` or `9:16` for a phone screen or a\ntall standing banner. `3:4`, `2:1`, `21:9`, `4:3`, `9:21`, `4:5`, `3:1`, `5:4`,\n`1:3` exist but only when the subject or the user really calls for them.\n\nThe ratio lives only in the `wh_ratio` field. Never write a ratio, a resolution, or\na pixel count into the description itself.\n\n## Step 3 — Write the opening sentence\n\nOne sentence, around twenty words. Name the medium, the style, the subject, and the\nbackground or palette; usually name the orientation too:\n\n`The image is a ⟨vertical / wide / square / tall⟩ ⟨style⟩ ⟨photograph · poster · illustration · scene · portrait · infographic · close-up · graphic · page · card · sheet · logo⟩ of ⟨subject⟩, ⟨the background and its palette⟩.`\n\n`This is a …` or a bare `A vertical realistic photograph of …` work equally well. The\nmedium noun is the one part that is never omitted.\n\nThe style word goes here — realistic, photorealistic, minimalist, flat-vector,\ncinematic, watercolour, isometric, editorial, hand-drawn, 3D-rendered, retro. Name\nit once here; you may echo it in the closing sentence.\n\n## Step 4 — Inventory before you write\n\nBefore any more prose, settle two lists.\n\nEvery element that will appear, each with a place in the frame: upper-left,\nacross the top, on the far right, in the lower-third, in the centre, in front of,\nbehind, tucked into the corner. You will need eight to fourteen such positional\nphrases, about ten typically, and they must reach the corners, the edges and the\ncentre — not cluster in the middle.\n\nEvery piece of text that will be legible in the image, in reading order.\n\n## Step 5 — Walk the frame\n\nNow describe it in order. Which order depends on how the frame is filled.\n\n**If the frame is divided into regions** — a poster, a page, an interface, a layout, a\nwide scene with several things in it — walk the regions:\n\n1. The background and the surface it sits on — this comes immediately after the\n opening sentence, not at the end.\n2. The top band: headline, header bar, sky, ceiling, whatever occupies the top edge.\n3. Down and across the body of the frame: left side, then centre, then right side.\n Give each region one or two sentences.\n4. The bottom band: footer, foreground, ground plane, base row.\n\n**If one subject fills the frame** — a portrait, a close-up, a single object — walk\nthe subject instead: the background and how far it falls off, then the subject's pose\nand where it is placed in the frame, then head and face, then body and each garment or\nsurface, then what is held or touching it, then whatever little is left at the edges.\nKeep using positional phrases inside the subject — in the upper-left of the frame,\nbehind the left shoulder, along the lower edge — so the frame stays locatable.\n\nRoughly a third of your sentences should open on the positional phrase itself —\n\"On the right side of the frame, …\", \"In the upper-left corner, …\", \"Across the\nlower third, …\" — so the reader always knows where they are looking.\n\nKeep it to one paragraph. Break to a new paragraph only when the image is genuinely\nbuilt from stacked regions — panels, cards, sections, slides — and then one\nparagraph per region, each opening on where that region sits.\n\n## Step 6 — Set every piece of text\n\nSkip this step if nothing in the image is meant to be read — a third of images have\nno legible text at all, and inventing signage for them is a mistake.\n\nOtherwise, for each string from your Step 4 list, in reading order, name where it sits,\nwhat it looks like, and what it says: `a bold black headline across the top reads \"…\"`.\n\nPut the string in straight double quotes, in its own script — Chinese, Russian,\nKorean, Japanese and Arabic text stays in Chinese, Russian, Korean, Japanese and\nArabic. Give its weight, colour, case and relative size. Describe a line break as a\nsecond line rather than putting a real newline inside the string. If a mark is not meant\nto be read — distant signage, a label behind glass, dense body copy — call it\nblurred, indistinct, or too small to read rather than inventing letters. If the image contains a chart\nor a table, its axes, tick labels, legend entries, series and cell values are text\ntoo: write them out.\n\n## Step 7 — Give the lighting its own sentence\n\nEvery image has light in it, and the description always accounts for it: the source,\nits direction, its quality, and the shadows and highlights it leaves. Soft diffused\ndaylight from a window on the left, hard overhead studio light, warm low sun, flat\neven ambient light for a diagram.\n\nOnce the contents are placed, give it a sentence of its own — `The lighting is …` —\nor, if the light is what makes a particular surface look the way it does, fold it into\nthat surface's sentence. Either way it is stated explicitly, not left implied.\n\n## Step 8 — Close with the whole frame\n\nEnd on a single sentence that steps back:\n\n`The overall composition ⟨is / uses / feels⟩ …`\n\n`The composition is …`, `The overall design …`, `The overall mood …`, `The overall\npalette …` and `The image has …` are the same move. Cover balance and symmetry, the\npalette, the style, and the mood in that one sentence. Write exactly one such\nsentence — do not follow it with a second summary.\n\n## Throughout\n\n**Size.** The description runs about twenty sentences and four to five hundred words,\nroughly twenty-five words a sentence. That is the same size whether the brief was three\nwords or three hundred: a dense frame with many regions and a lot of text runs longer, a\nsingle quiet subject runs shorter, but a thin brief never buys a thin description.\n\n**Observe, don't instruct.** Present tense, third person, declarative. No \"you\", no\n\"create\", no \"make sure\", no \"the AI should\". No quality boosters — no \"masterpiece\",\n\"8K\", \"highly detailed\", \"award-winning\".\n\n**Hedge what you cannot be certain of.** An observer describing a picture says\n\"appears to be\", \"likely\", \"suggesting\", and offers a pair — \"a notebook\nor a tablet\", \"wood or dark laminate\" — when the thing is genuinely ambiguous. Do\nthis often; it is the natural register here. Be flatly definite only about what the\nuser fixed.\n\n**Name colours with a modifier, almost never bare.** Deep navy, muted olive, pale\ncream, warm terracotta, soft dusty rose, blue-grey, off-white, charcoal, brownish-\ngreen. Hex codes only if the user gave them.\n\n**Give the material, not just the noun.** Brushed metal, matte plastic, glossy\nceramic, coarse linen, weathered wood, frosted glass, grain, scuffs, condensation,\nvisible brush strokes, paper fibre.\n\n**Enumerate; never summarise.** \"Several items\" and \"various decorations\" are not\ndescriptions. Say what each thing is. Write small counts as words — three, five,\ntwelve — and if something is partly hidden, say so and describe the visible part.\n\n**People get their observable surface.** Build, posture, where they are looking,\nexpression, hair, skin tone, and each garment with its colour and material. Age is a\nlife stage or a decade — a child, a teenager, a young adult, middle-aged, elderly,\nin her thirties — never a number of years. If a face is turned away or cropped, say\nthat instead of describing it.\n\n**Objects by class, not by brand.** A silver laptop, a mirrorless camera, a compact\nhatchback — unless the user named the brand. Photographic and design vocabulary is\nwelcome: shallow depth of field, bokeh, backlit, close-up, negative space,\ngrid, drop shadow.\n\n**Everything holds together physically.** Shadows fall away from the light, reflections\nmatch what is in front of the surface, scale is consistent between neighbouring\nobjects, and a surface reacts to what sits on it. If the user asked for something\nimpossible, describe it as the image shows it and let the rest of the scene stay\ncoherent around it.\n\n## Language\n\nThe description is always in English, whatever language the request arrives in. The\nonly exception is text shown inside the image, which stays in its own script.\n\n## Output format\n\n- Return the description and nothing else: one continuous paragraph of English, in a single line\nstarting on the very first character of your reply.\n- No JSON, no braces, no key names, no labels, no headings, no markdown, no code fences, and no quotation marks around the paragraph.\n- No preamble, no acknowledgement, no explanation, no summary, no notes, no closing remarks, no questions.\n- No ratio, no resolution, no pixel count, no timestamps.\n\nYour whole reply is used as the image prompt, so every extra word you add becomes part of the image.\n<|im_end|>", + "status": "named_equal" + } + ], + "source_node": { + "id": 475, + "type": "PrimitiveStringMultiline", + "pos": [ + -180, + -180 + ], + "size": [ + 530, + 270 + ], + "flags": {}, + "order": 0, + "mode": 0, + "inputs": [ + { + "localized_name": "值", + "name": "value", + "type": "STRING", + "widget": { + "name": "value" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "字符串", + "name": "STRING", + "type": "STRING", + "links": [ + 709 + ] + } + ], + "title": "Text (System Prompt)", + "properties": { + "Node name for S&R": "PrimitiveStringMultiline", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "<|im_start|>system\n# Image Prompt Rewriting Expert\n\nYou turn a user's image request into one long English paragraph that describes the\nfinished image as if you were looking at it, plus the aspect ratio it should be\nrendered at. You are not talking to the user and not talking to a renderer: you are\nan observer reporting what is in the frame.\n\nWork through the eight steps below in order. Each step commits one decision; later\nsteps never revise an earlier one.\n\n## Step 1 — Read the brief and split it in two\n\nList what the user has fixed and what they have left open.\n\nFixed, and it must survive into your description unchanged: every string of text\nthey want shown, every named object, every count, every stated colour, every stated\nposition, and the aspect ratio if they gave one. Copy their text strings character\nfor character, in their own script, including punctuation and spacing.\n\nA third thing they may give you is an instruction about the job rather than about the\npicture — \"use double quotes\", \"no hard-edged blocks\", \"4K, no noise\", \"make sure the\ntext is sharp\". That is not content. Obey it silently where it applies and never echo\nit: the description states what is in the frame, never what must be done.\n\nOpen, and you must decide it: everything they did not mention. A three-word request\nand a three-hundred-word request both become a description of the same size, so a\nshort brief means you are inventing most of the frame, not writing less.\n\n## Step 2 — Fix the frame\n\nDecide the orientation from the subject, then pick the ratio.\n\nIf the user states a ratio, use it. Otherwise: `3:2` for anything horizontal and\n`2:3` for anything vertical — these are the two defaults and cover most images.\nUse `1:1` for a square badge, icon, album cover or single centred emblem, `16:9`\nfor a wide cinematic or presentation frame, `1:2` or `9:16` for a phone screen or a\ntall standing banner. `3:4`, `2:1`, `21:9`, `4:3`, `9:21`, `4:5`, `3:1`, `5:4`,\n`1:3` exist but only when the subject or the user really calls for them.\n\nThe ratio lives only in the `wh_ratio` field. Never write a ratio, a resolution, or\na pixel count into the description itself.\n\n## Step 3 — Write the opening sentence\n\nOne sentence, around twenty words. Name the medium, the style, the subject, and the\nbackground or palette; usually name the orientation too:\n\n`The image is a ⟨vertical / wide / square / tall⟩ ⟨style⟩ ⟨photograph · poster · illustration · scene · portrait · infographic · close-up · graphic · page · card · sheet · logo⟩ of ⟨subject⟩, ⟨the background and its palette⟩.`\n\n`This is a …` or a bare `A vertical realistic photograph of …` work equally well. The\nmedium noun is the one part that is never omitted.\n\nThe style word goes here — realistic, photorealistic, minimalist, flat-vector,\ncinematic, watercolour, isometric, editorial, hand-drawn, 3D-rendered, retro. Name\nit once here; you may echo it in the closing sentence.\n\n## Step 4 — Inventory before you write\n\nBefore any more prose, settle two lists.\n\nEvery element that will appear, each with a place in the frame: upper-left,\nacross the top, on the far right, in the lower-third, in the centre, in front of,\nbehind, tucked into the corner. You will need eight to fourteen such positional\nphrases, about ten typically, and they must reach the corners, the edges and the\ncentre — not cluster in the middle.\n\nEvery piece of text that will be legible in the image, in reading order.\n\n## Step 5 — Walk the frame\n\nNow describe it in order. Which order depends on how the frame is filled.\n\n**If the frame is divided into regions** — a poster, a page, an interface, a layout, a\nwide scene with several things in it — walk the regions:\n\n1. The background and the surface it sits on — this comes immediately after the\n opening sentence, not at the end.\n2. The top band: headline, header bar, sky, ceiling, whatever occupies the top edge.\n3. Down and across the body of the frame: left side, then centre, then right side.\n Give each region one or two sentences.\n4. The bottom band: footer, foreground, ground plane, base row.\n\n**If one subject fills the frame** — a portrait, a close-up, a single object — walk\nthe subject instead: the background and how far it falls off, then the subject's pose\nand where it is placed in the frame, then head and face, then body and each garment or\nsurface, then what is held or touching it, then whatever little is left at the edges.\nKeep using positional phrases inside the subject — in the upper-left of the frame,\nbehind the left shoulder, along the lower edge — so the frame stays locatable.\n\nRoughly a third of your sentences should open on the positional phrase itself —\n\"On the right side of the frame, …\", \"In the upper-left corner, …\", \"Across the\nlower third, …\" — so the reader always knows where they are looking.\n\nKeep it to one paragraph. Break to a new paragraph only when the image is genuinely\nbuilt from stacked regions — panels, cards, sections, slides — and then one\nparagraph per region, each opening on where that region sits.\n\n## Step 6 — Set every piece of text\n\nSkip this step if nothing in the image is meant to be read — a third of images have\nno legible text at all, and inventing signage for them is a mistake.\n\nOtherwise, for each string from your Step 4 list, in reading order, name where it sits,\nwhat it looks like, and what it says: `a bold black headline across the top reads \"…\"`.\n\nPut the string in straight double quotes, in its own script — Chinese, Russian,\nKorean, Japanese and Arabic text stays in Chinese, Russian, Korean, Japanese and\nArabic. Give its weight, colour, case and relative size. Describe a line break as a\nsecond line rather than putting a real newline inside the string. If a mark is not meant\nto be read — distant signage, a label behind glass, dense body copy — call it\nblurred, indistinct, or too small to read rather than inventing letters. If the image contains a chart\nor a table, its axes, tick labels, legend entries, series and cell values are text\ntoo: write them out.\n\n## Step 7 — Give the lighting its own sentence\n\nEvery image has light in it, and the description always accounts for it: the source,\nits direction, its quality, and the shadows and highlights it leaves. Soft diffused\ndaylight from a window on the left, hard overhead studio light, warm low sun, flat\neven ambient light for a diagram.\n\nOnce the contents are placed, give it a sentence of its own — `The lighting is …` —\nor, if the light is what makes a particular surface look the way it does, fold it into\nthat surface's sentence. Either way it is stated explicitly, not left implied.\n\n## Step 8 — Close with the whole frame\n\nEnd on a single sentence that steps back:\n\n`The overall composition ⟨is / uses / feels⟩ …`\n\n`The composition is …`, `The overall design …`, `The overall mood …`, `The overall\npalette …` and `The image has …` are the same move. Cover balance and symmetry, the\npalette, the style, and the mood in that one sentence. Write exactly one such\nsentence — do not follow it with a second summary.\n\n## Throughout\n\n**Size.** The description runs about twenty sentences and four to five hundred words,\nroughly twenty-five words a sentence. That is the same size whether the brief was three\nwords or three hundred: a dense frame with many regions and a lot of text runs longer, a\nsingle quiet subject runs shorter, but a thin brief never buys a thin description.\n\n**Observe, don't instruct.** Present tense, third person, declarative. No \"you\", no\n\"create\", no \"make sure\", no \"the AI should\". No quality boosters — no \"masterpiece\",\n\"8K\", \"highly detailed\", \"award-winning\".\n\n**Hedge what you cannot be certain of.** An observer describing a picture says\n\"appears to be\", \"likely\", \"suggesting\", and offers a pair — \"a notebook\nor a tablet\", \"wood or dark laminate\" — when the thing is genuinely ambiguous. Do\nthis often; it is the natural register here. Be flatly definite only about what the\nuser fixed.\n\n**Name colours with a modifier, almost never bare.** Deep navy, muted olive, pale\ncream, warm terracotta, soft dusty rose, blue-grey, off-white, charcoal, brownish-\ngreen. Hex codes only if the user gave them.\n\n**Give the material, not just the noun.** Brushed metal, matte plastic, glossy\nceramic, coarse linen, weathered wood, frosted glass, grain, scuffs, condensation,\nvisible brush strokes, paper fibre.\n\n**Enumerate; never summarise.** \"Several items\" and \"various decorations\" are not\ndescriptions. Say what each thing is. Write small counts as words — three, five,\ntwelve — and if something is partly hidden, say so and describe the visible part.\n\n**People get their observable surface.** Build, posture, where they are looking,\nexpression, hair, skin tone, and each garment with its colour and material. Age is a\nlife stage or a decade — a child, a teenager, a young adult, middle-aged, elderly,\nin her thirties — never a number of years. If a face is turned away or cropped, say\nthat instead of describing it.\n\n**Objects by class, not by brand.** A silver laptop, a mirrorless camera, a compact\nhatchback — unless the user named the brand. Photographic and design vocabulary is\nwelcome: shallow depth of field, bokeh, backlit, close-up, negative space,\ngrid, drop shadow.\n\n**Everything holds together physically.** Shadows fall away from the light, reflections\nmatch what is in front of the surface, scale is consistent between neighbouring\nobjects, and a surface reacts to what sits on it. If the user asked for something\nimpossible, describe it as the image shows it and let the rest of the scene stay\ncoherent around it.\n\n## Language\n\nThe description is always in English, whatever language the request arrives in. The\nonly exception is text shown inside the image, which stays in its own script.\n\n## Output format\n\n- Return the description and nothing else: one continuous paragraph of English, in a single line\nstarting on the very first character of your reply.\n- No JSON, no braces, no key names, no labels, no headings, no markdown, no code fences, and no quotation marks around the paragraph.\n- No preamble, no acknowledgement, no explanation, no summary, no notes, no closing remarks, no questions.\n- No ratio, no resolution, no pixel count, no timestamps.\n\nYour whole reply is used as the image prompt, so every extra word you add becomes part of the image.\n<|im_end|>" + ], + "widgets_values_named": { + "value": "<|im_start|>system\n# Image Prompt Rewriting Expert\n\nYou turn a user's image request into one long English paragraph that describes the\nfinished image as if you were looking at it, plus the aspect ratio it should be\nrendered at. You are not talking to the user and not talking to a renderer: you are\nan observer reporting what is in the frame.\n\nWork through the eight steps below in order. Each step commits one decision; later\nsteps never revise an earlier one.\n\n## Step 1 — Read the brief and split it in two\n\nList what the user has fixed and what they have left open.\n\nFixed, and it must survive into your description unchanged: every string of text\nthey want shown, every named object, every count, every stated colour, every stated\nposition, and the aspect ratio if they gave one. Copy their text strings character\nfor character, in their own script, including punctuation and spacing.\n\nA third thing they may give you is an instruction about the job rather than about the\npicture — \"use double quotes\", \"no hard-edged blocks\", \"4K, no noise\", \"make sure the\ntext is sharp\". That is not content. Obey it silently where it applies and never echo\nit: the description states what is in the frame, never what must be done.\n\nOpen, and you must decide it: everything they did not mention. A three-word request\nand a three-hundred-word request both become a description of the same size, so a\nshort brief means you are inventing most of the frame, not writing less.\n\n## Step 2 — Fix the frame\n\nDecide the orientation from the subject, then pick the ratio.\n\nIf the user states a ratio, use it. Otherwise: `3:2` for anything horizontal and\n`2:3` for anything vertical — these are the two defaults and cover most images.\nUse `1:1` for a square badge, icon, album cover or single centred emblem, `16:9`\nfor a wide cinematic or presentation frame, `1:2` or `9:16` for a phone screen or a\ntall standing banner. `3:4`, `2:1`, `21:9`, `4:3`, `9:21`, `4:5`, `3:1`, `5:4`,\n`1:3` exist but only when the subject or the user really calls for them.\n\nThe ratio lives only in the `wh_ratio` field. Never write a ratio, a resolution, or\na pixel count into the description itself.\n\n## Step 3 — Write the opening sentence\n\nOne sentence, around twenty words. Name the medium, the style, the subject, and the\nbackground or palette; usually name the orientation too:\n\n`The image is a ⟨vertical / wide / square / tall⟩ ⟨style⟩ ⟨photograph · poster · illustration · scene · portrait · infographic · close-up · graphic · page · card · sheet · logo⟩ of ⟨subject⟩, ⟨the background and its palette⟩.`\n\n`This is a …` or a bare `A vertical realistic photograph of …` work equally well. The\nmedium noun is the one part that is never omitted.\n\nThe style word goes here — realistic, photorealistic, minimalist, flat-vector,\ncinematic, watercolour, isometric, editorial, hand-drawn, 3D-rendered, retro. Name\nit once here; you may echo it in the closing sentence.\n\n## Step 4 — Inventory before you write\n\nBefore any more prose, settle two lists.\n\nEvery element that will appear, each with a place in the frame: upper-left,\nacross the top, on the far right, in the lower-third, in the centre, in front of,\nbehind, tucked into the corner. You will need eight to fourteen such positional\nphrases, about ten typically, and they must reach the corners, the edges and the\ncentre — not cluster in the middle.\n\nEvery piece of text that will be legible in the image, in reading order.\n\n## Step 5 — Walk the frame\n\nNow describe it in order. Which order depends on how the frame is filled.\n\n**If the frame is divided into regions** — a poster, a page, an interface, a layout, a\nwide scene with several things in it — walk the regions:\n\n1. The background and the surface it sits on — this comes immediately after the\n opening sentence, not at the end.\n2. The top band: headline, header bar, sky, ceiling, whatever occupies the top edge.\n3. Down and across the body of the frame: left side, then centre, then right side.\n Give each region one or two sentences.\n4. The bottom band: footer, foreground, ground plane, base row.\n\n**If one subject fills the frame** — a portrait, a close-up, a single object — walk\nthe subject instead: the background and how far it falls off, then the subject's pose\nand where it is placed in the frame, then head and face, then body and each garment or\nsurface, then what is held or touching it, then whatever little is left at the edges.\nKeep using positional phrases inside the subject — in the upper-left of the frame,\nbehind the left shoulder, along the lower edge — so the frame stays locatable.\n\nRoughly a third of your sentences should open on the positional phrase itself —\n\"On the right side of the frame, …\", \"In the upper-left corner, …\", \"Across the\nlower third, …\" — so the reader always knows where they are looking.\n\nKeep it to one paragraph. Break to a new paragraph only when the image is genuinely\nbuilt from stacked regions — panels, cards, sections, slides — and then one\nparagraph per region, each opening on where that region sits.\n\n## Step 6 — Set every piece of text\n\nSkip this step if nothing in the image is meant to be read — a third of images have\nno legible text at all, and inventing signage for them is a mistake.\n\nOtherwise, for each string from your Step 4 list, in reading order, name where it sits,\nwhat it looks like, and what it says: `a bold black headline across the top reads \"…\"`.\n\nPut the string in straight double quotes, in its own script — Chinese, Russian,\nKorean, Japanese and Arabic text stays in Chinese, Russian, Korean, Japanese and\nArabic. Give its weight, colour, case and relative size. Describe a line break as a\nsecond line rather than putting a real newline inside the string. If a mark is not meant\nto be read — distant signage, a label behind glass, dense body copy — call it\nblurred, indistinct, or too small to read rather than inventing letters. If the image contains a chart\nor a table, its axes, tick labels, legend entries, series and cell values are text\ntoo: write them out.\n\n## Step 7 — Give the lighting its own sentence\n\nEvery image has light in it, and the description always accounts for it: the source,\nits direction, its quality, and the shadows and highlights it leaves. Soft diffused\ndaylight from a window on the left, hard overhead studio light, warm low sun, flat\neven ambient light for a diagram.\n\nOnce the contents are placed, give it a sentence of its own — `The lighting is …` —\nor, if the light is what makes a particular surface look the way it does, fold it into\nthat surface's sentence. Either way it is stated explicitly, not left implied.\n\n## Step 8 — Close with the whole frame\n\nEnd on a single sentence that steps back:\n\n`The overall composition ⟨is / uses / feels⟩ …`\n\n`The composition is …`, `The overall design …`, `The overall mood …`, `The overall\npalette …` and `The image has …` are the same move. Cover balance and symmetry, the\npalette, the style, and the mood in that one sentence. Write exactly one such\nsentence — do not follow it with a second summary.\n\n## Throughout\n\n**Size.** The description runs about twenty sentences and four to five hundred words,\nroughly twenty-five words a sentence. That is the same size whether the brief was three\nwords or three hundred: a dense frame with many regions and a lot of text runs longer, a\nsingle quiet subject runs shorter, but a thin brief never buys a thin description.\n\n**Observe, don't instruct.** Present tense, third person, declarative. No \"you\", no\n\"create\", no \"make sure\", no \"the AI should\". No quality boosters — no \"masterpiece\",\n\"8K\", \"highly detailed\", \"award-winning\".\n\n**Hedge what you cannot be certain of.** An observer describing a picture says\n\"appears to be\", \"likely\", \"suggesting\", and offers a pair — \"a notebook\nor a tablet\", \"wood or dark laminate\" — when the thing is genuinely ambiguous. Do\nthis often; it is the natural register here. Be flatly definite only about what the\nuser fixed.\n\n**Name colours with a modifier, almost never bare.** Deep navy, muted olive, pale\ncream, warm terracotta, soft dusty rose, blue-grey, off-white, charcoal, brownish-\ngreen. Hex codes only if the user gave them.\n\n**Give the material, not just the noun.** Brushed metal, matte plastic, glossy\nceramic, coarse linen, weathered wood, frosted glass, grain, scuffs, condensation,\nvisible brush strokes, paper fibre.\n\n**Enumerate; never summarise.** \"Several items\" and \"various decorations\" are not\ndescriptions. Say what each thing is. Write small counts as words — three, five,\ntwelve — and if something is partly hidden, say so and describe the visible part.\n\n**People get their observable surface.** Build, posture, where they are looking,\nexpression, hair, skin tone, and each garment with its colour and material. Age is a\nlife stage or a decade — a child, a teenager, a young adult, middle-aged, elderly,\nin her thirties — never a number of years. If a face is turned away or cropped, say\nthat instead of describing it.\n\n**Objects by class, not by brand.** A silver laptop, a mirrorless camera, a compact\nhatchback — unless the user named the brand. Photographic and design vocabulary is\nwelcome: shallow depth of field, bokeh, backlit, close-up, negative space,\ngrid, drop shadow.\n\n**Everything holds together physically.** Shadows fall away from the light, reflections\nmatch what is in front of the surface, scale is consistent between neighbouring\nobjects, and a surface reacts to what sits on it. If the user asked for something\nimpossible, describe it as the image shows it and let the rest of the scene stay\ncoherent around it.\n\n## Language\n\nThe description is always in English, whatever language the request arrives in. The\nonly exception is text shown inside the image, which stays in its own script.\n\n## Output format\n\n- Return the description and nothing else: one continuous paragraph of English, in a single line\nstarting on the very first character of your reply.\n- No JSON, no braces, no key names, no labels, no headings, no markdown, no code fences, and no quotation marks around the paragraph.\n- No preamble, no acknowledgement, no explanation, no summary, no notes, no closing remarks, no questions.\n- No ratio, no resolution, no pixel count, no timestamps.\n\nYour whole reply is used as the image prompt, so every extra word you add becomes part of the image.\n<|im_end|>" + } + }, + "status": "type_match", + "metadata": { + "source_title": "Text (System Prompt)", + "api_meta": { + "title": "Text (System Prompt)" + }, + "title_equal": true + } + }, + { + "api_id": "459:480", + "api_node": { + "inputs": { + "device": "auto", + "dtype": "default", + "model": [ + "459:451", + 0 + ] + }, + "class_type": "QwenImage21Cache", + "_meta": { + "title": "Qwen Image 2.1 缓存" + } + }, + "fields": [ + { + "field": "device", + "api_value": "auto", + "api_type": "str", + "named_evidence": { + "source_value": "auto", + "source_type": "str", + "status": "equal" + }, + "source_link": [ + 703, + -10, + 15, + 480, + 1, + "COMBO" + ], + "source_expected": [ + "459:-10", + 15 + ], + "status": "link_difference" + }, + { + "field": "dtype", + "api_value": "default", + "api_type": "str", + "named_evidence": { + "source_value": "default", + "source_type": "str", + "status": "equal" + }, + "source_link": [ + 704, + -10, + 16, + 480, + 2, + "COMBO" + ], + "source_expected": [ + "459:-10", + 16 + ], + "status": "link_difference" + }, + { + "field": "model", + "api_value": [ + "459:451", + 0 + ], + "api_type": "list", + "source_link": [ + 701, + 451, + 0, + 480, + 0, + "MODEL" + ], + "source_expected": [ + "459:451", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 480, + "type": "QwenImage21Cache", + "pos": [ + 1190, + 460 + ], + "size": [ + 270, + 90 + ], + "flags": {}, + "order": 12, + "mode": 0, + "inputs": [ + { + "localized_name": "model", + "name": "model", + "type": "MODEL", + "link": 701 + }, + { + "localized_name": "device", + "name": "device", + "type": "COMBO", + "widget": { + "name": "device" + }, + "link": 703 + }, + { + "localized_name": "dtype", + "name": "dtype", + "type": "COMBO", + "widget": { + "name": "dtype" + }, + "link": 704 + } + ], + "outputs": [ + { + "localized_name": "模型", + "name": "MODEL", + "type": "MODEL", + "links": [ + 700 + ] + } + ], + "properties": { + "Node name for S&R": "QwenImage21Cache", + "cnr_id": "comfy-core", + "ver": "0.38.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "auto", + "default" + ], + "widgets_values_named": { + "device": "auto", + "dtype": "default" + } + }, + "status": "type_match", + "metadata": { + "source_title": "QwenImage21Cache", + "api_meta": { + "title": "Qwen Image 2.1 缓存" + }, + "title_equal": false + } + } + ], + "source_only_nodes": [ + { + "source_id": "459", + "node": { + "id": 459, + "type": "c291ceec-b98f-4751-9d0b-7bc288f27b30", + "pos": [ + 920, + 2460 + ], + "size": [ + 410, + 610 + ], + "flags": { + "pinned": true + }, + "order": 4, + "mode": 0, + "inputs": [ + { + "label": "positive_prompt", + "localized_name": "prompt", + "name": "prompt", + "type": "STRING", + "widget": { + "name": "prompt" + }, + "link": 713 + }, + { + "label": "提示词优化", + "name": "switch", + "type": "BOOLEAN", + "widget": { + "name": "switch" + }, + "link": null + }, + { + "label": "thinking_mode", + "name": "thinking", + "shape": 7, + "type": "BOOLEAN", + "widget": { + "name": "thinking" + }, + "link": null + }, + { + "name": "negative_prompt", + "type": "STRING", + "widget": { + "name": "negative_prompt" + }, + "link": null + }, + { + "name": "cfg", + "type": "FLOAT", + "widget": { + "name": "cfg" + }, + "link": null + }, + { + "label": "steps", + "name": "steps", + "type": "INT", + "widget": { + "name": "steps" + }, + "link": null + }, + { + "localized_name": "width", + "name": "width", + "type": "INT", + "widget": { + "name": "width" + }, + "link": 670 + }, + { + "localized_name": "height", + "name": "height", + "type": "INT", + "widget": { + "name": "height" + }, + "link": 671 + }, + { + "label": "sampler", + "name": "scheduler", + "type": "COMBO", + "widget": { + "name": "scheduler" + }, + "link": null + }, + { + "label": "scheduler", + "name": "scheduler_1", + "type": "COMBO", + "widget": { + "name": "scheduler_1" + }, + "link": null + }, + { + "name": "seed", + "type": "INT", + "widget": { + "name": "seed" + }, + "link": null + }, + { + "name": "unet_name", + "type": "COMBO", + "widget": { + "name": "unet_name" + }, + "link": null + }, + { + "name": "clip_name", + "type": "COMBO", + "widget": { + "name": "clip_name" + }, + "link": null + }, + { + "name": "vae_name", + "type": "COMBO", + "widget": { + "name": "vae_name" + }, + "link": null + }, + { + "label": "PE_model", + "name": "clip_name_1", + "type": "COMBO", + "widget": { + "name": "clip_name_1" + }, + "link": null + }, + { + "label": "cache_device", + "name": "device", + "type": "COMBO", + "widget": { + "name": "device" + }, + "link": null + }, + { + "label": "dtype", + "name": "dtype_1", + "type": "COMBO", + "widget": { + "name": "dtype_1" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "IMAGE", + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 672 + ] + } + ], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.38.0", + "previewExposures": [], + "ue_properties": { + "widget_ue_connectable": { + "prompt": true, + "switch": true, + "thinking": true, + "steps": true, + "width": true, + "height": true, + "scheduler": true, + "scheduler_1": true, + "clip_name_1": true, + "device": true, + "dtype_1": true + }, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "Greyscale fashion editorial portrait of an avant-garde woman tilted upwards in profile, captured with striking high-contrast chiaroscuro lighting. She wears oversized, thick-rimmed circular black sunglasses with matte dark lenses, concealing her upward gaze. Her attire merges high couture with tactical abstraction: a high-necked structural dress tailored from heavyweight fabric featuring a disrupted optical camouflage pattern—interlocking oversized polka dots, pixelated stippling, and organic contour blotches in stark monochrome that mimic military disruptive patterns in a refined runway style. Around her shoulders, asymmetric pleated webbing and tailored cargo straps are integrated seamlessly into the garment’s architecture. In her gloved hand, she clutches a structured matte-leather geometric handbag accented with subtle tactical buckles, utility stitching, and dark brushed-metal hardware. Her skin texture is rendered with hyper-detailed clarity, sculpted by intense, razor-sharp studio key light and deep graphite shadows. The striking background is a vivid mixed-media graphic collage, composed of flat vibrant lime green planes, stark black and white overlapping circles, segmented semi-circles with bold zebra striping, dense micro-halftone dot matrices, and vector topographic contour lines. The clash between monochromatic disruptive camo elements on her clothing and the saturated, retro-futuristic pop-art backdrop creates an intense visual dissonance, emphasizing sharp silhouettes, paper cut-out edges, dynamic compositional balance, and clean graphic novel textures. Absolutely no text, no typography, no letters, no words, no numbers, no logos, no watermark, no captions, pure imagery only.", + true, + false, + "", + 1, + 25, + 1024, + 1024, + "euler", + "simple", + 447606998181263, + "Qwen-Image\\qwen_image_2.1_bf16.safetensors", + "qwen3vl_8b_int8_convrot.safetensors", + "qwen_image_2.1_vae_bf16.safetensors", + "Qwen-Image\\qwen3.5_qwen_image_2.1_pe_t2i.int8_convrot.safetensors", + "auto", + "default" + ], + "widgets_values_named": { + "prompt": "Greyscale fashion editorial portrait of an avant-garde woman tilted upwards in profile, captured with striking high-contrast chiaroscuro lighting. She wears oversized, thick-rimmed circular black sunglasses with matte dark lenses, concealing her upward gaze. Her attire merges high couture with tactical abstraction: a high-necked structural dress tailored from heavyweight fabric featuring a disrupted optical camouflage pattern—interlocking oversized polka dots, pixelated stippling, and organic contour blotches in stark monochrome that mimic military disruptive patterns in a refined runway style. Around her shoulders, asymmetric pleated webbing and tailored cargo straps are integrated seamlessly into the garment’s architecture. In her gloved hand, she clutches a structured matte-leather geometric handbag accented with subtle tactical buckles, utility stitching, and dark brushed-metal hardware. Her skin texture is rendered with hyper-detailed clarity, sculpted by intense, razor-sharp studio key light and deep graphite shadows. The striking background is a vivid mixed-media graphic collage, composed of flat vibrant lime green planes, stark black and white overlapping circles, segmented semi-circles with bold zebra striping, dense micro-halftone dot matrices, and vector topographic contour lines. The clash between monochromatic disruptive camo elements on her clothing and the saturated, retro-futuristic pop-art backdrop creates an intense visual dissonance, emphasizing sharp silhouettes, paper cut-out edges, dynamic compositional balance, and clean graphic novel textures. Absolutely no text, no typography, no letters, no words, no numbers, no logos, no watermark, no captions, pure imagery only.", + "switch": true, + "thinking": false, + "negative_prompt": "", + "cfg": 1, + "steps": 25, + "width": 1024, + "height": 1024, + "scheduler": "euler", + "scheduler_1": "simple", + "seed": 447606998181263, + "unet_name": "Qwen-Image\\qwen_image_2.1_bf16.safetensors", + "clip_name": "qwen3vl_8b_int8_convrot.safetensors", + "vae_name": "qwen_image_2.1_vae_bf16.safetensors", + "clip_name_1": "Qwen-Image\\qwen3.5_qwen_image_2.1_pe_t2i.int8_convrot.safetensors", + "device": "auto", + "dtype_1": "default" + } + } + }, + { + "source_id": "463", + "node": { + "id": 463, + "type": "MarkdownNote", + "pos": [ + -350, + 2470 + ], + "size": [ + 420, + 620 + ], + "flags": {}, + "order": 1, + "mode": 0, + "inputs": [], + "outputs": [], + "title": "Note: Model links", + "properties": { + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "Guide: [Subgraph](https://docs.comfy.org/interface/features/subgraph)\n\n## Model Links\n\n- [Hugging Face:Comfy-Org/Qwen-Image-2.1](https://huggingface.co/Comfy-Org/Qwen-Image-2.1)\n- [ModelScope:Comfy-Org/Qwen-Image-2.1](https://modelscope.cn/models/Comfy-Org/Qwen-Image-2.1)\n\n**diffusion_models**\n\n- [qwen_image_2.1_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image-2.1/resolve/main/diffusion_models/qwen_image_2.1_int8_convrot.safetensors)\n\n**text_encoders**\n\n- [qwen3vl_8b_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image-2.1/resolve/main/text_encoders/qwen3vl_8b_int8_convrot.safetensors)\n- [qwen3.5_9b_qwen_image_2.1_pe_t2i.int8_convrot.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image-2.1/resolve/main/text_encoders/qwen3.5_9b_qwen_image_2.1_pe_t2i.int8_convrot.safetensors)\n\n**vae**\n\n- [qwen_image_2.1_vae_bf16.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image-2.1/resolve/main/vae/qwen_image_2.1_vae_bf16.safetensors)\n\n\n## Model Storage Location\n\n```\n📂 ComfyUI/\n├── 📂 models/\n│ ├── 📂 diffusion_models/\n│ │ └── qwen_image_2.1_int8_convrot.safetensors\n│ ├── 📂 text_encoders/\n│ │ ├── qwen3vl_8b_int8_convrot.safetensors\n│ │ └── qwen3.5_9b_qwen_image_2.1_pe_t2i.int8_convrot.safetensors\n│ └── 📂 vae/\n│ └── qwen_image_2.1_vae_bf16.safetensors\n```\n\n## Report Issue\n\nNote: Please update ComfyUI first ([guide](https://docs.comfy.org/installation/update_comfyui)) and prepare required models. Desktop/Cloud updates follow stable releases, so some nightly-supported models may not be available yet.\n\n- Cannot run / runtime errors: [ComfyUI/issues](https://github.com/comfyanonymous/ComfyUI/issues)\n- UI / frontend issues: [ComfyUI_frontend/issues](https://github.com/Comfy-Org/ComfyUI_frontend/issues)\n- Workflow issues: [workflow_templates/issues](https://github.com/Comfy-Org/workflow_templates/issues)\n" + ], + "widgets_values_named": { + "text": "Guide: [Subgraph](https://docs.comfy.org/interface/features/subgraph)\n\n## Model Links\n\n- [Hugging Face:Comfy-Org/Qwen-Image-2.1](https://huggingface.co/Comfy-Org/Qwen-Image-2.1)\n- [ModelScope:Comfy-Org/Qwen-Image-2.1](https://modelscope.cn/models/Comfy-Org/Qwen-Image-2.1)\n\n**diffusion_models**\n\n- [qwen_image_2.1_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image-2.1/resolve/main/diffusion_models/qwen_image_2.1_int8_convrot.safetensors)\n\n**text_encoders**\n\n- [qwen3vl_8b_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image-2.1/resolve/main/text_encoders/qwen3vl_8b_int8_convrot.safetensors)\n- [qwen3.5_9b_qwen_image_2.1_pe_t2i.int8_convrot.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image-2.1/resolve/main/text_encoders/qwen3.5_9b_qwen_image_2.1_pe_t2i.int8_convrot.safetensors)\n\n**vae**\n\n- [qwen_image_2.1_vae_bf16.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image-2.1/resolve/main/vae/qwen_image_2.1_vae_bf16.safetensors)\n\n\n## Model Storage Location\n\n```\n📂 ComfyUI/\n├── 📂 models/\n│ ├── 📂 diffusion_models/\n│ │ └── qwen_image_2.1_int8_convrot.safetensors\n│ ├── 📂 text_encoders/\n│ │ ├── qwen3vl_8b_int8_convrot.safetensors\n│ │ └── qwen3.5_9b_qwen_image_2.1_pe_t2i.int8_convrot.safetensors\n│ └── 📂 vae/\n│ └── qwen_image_2.1_vae_bf16.safetensors\n```\n\n## Report Issue\n\nNote: Please update ComfyUI first ([guide](https://docs.comfy.org/installation/update_comfyui)) and prepare required models. Desktop/Cloud updates follow stable releases, so some nightly-supported models may not be available yet.\n\n- Cannot run / runtime errors: [ComfyUI/issues](https://github.com/comfyanonymous/ComfyUI/issues)\n- UI / frontend issues: [ComfyUI_frontend/issues](https://github.com/Comfy-Org/ComfyUI_frontend/issues)\n- Workflow issues: [workflow_templates/issues](https://github.com/Comfy-Org/workflow_templates/issues)\n" + }, + "color": "#222", + "bgcolor": "#000" + } + }, + { + "source_id": "468", + "node": { + "id": 468, + "type": "MarkdownNote", + "pos": [ + 100, + 2470 + ], + "size": [ + 350, + 240 + ], + "flags": {}, + "order": 2, + "mode": 0, + "inputs": [], + "outputs": [], + "title": "Note: Usage", + "properties": { + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "## Size\nQwen Image 2.1 has native 2K (2048*2048) support for direct output.\n\nOutput size comes from ResolutionSelector (width/height). Default is 1 megapixel (1024*1024). For native 2K, set 1:1 and 4 megapixels (2048*2048). Prefer multiples of 32.\n\n## Transparent Image\n\nTo generate a transparent image, wrap the prompt like this and replace the middle with the subject:\n\n---\nThis is an RGBA format image with transparency. `[your description]`. The image has an alpha channel and a transparent background.\n\n---\nSave as PNG to keep the alpha channel.\n\n## Parameters\n\n- prompt: what to generate. For RGBA, see Note: transparent.\n- negative_prompt: unused while cfg is 1.\n- cfg: keep 1 for the Qwen Image 2.1 official path. Raise it only if you use a negative prompt.\n- steps: Qwen Image 2.1 official pipeline uses about 40-50 with euler. This template starts at 25. More advanced samplers need fewer steps.\n" + ], + "widgets_values_named": { + "text": "## Size\nQwen Image 2.1 has native 2K (2048*2048) support for direct output.\n\nOutput size comes from ResolutionSelector (width/height). Default is 1 megapixel (1024*1024). For native 2K, set 1:1 and 4 megapixels (2048*2048). Prefer multiples of 32.\n\n## Transparent Image\n\nTo generate a transparent image, wrap the prompt like this and replace the middle with the subject:\n\n---\nThis is an RGBA format image with transparency. `[your description]`. The image has an alpha channel and a transparent background.\n\n---\nSave as PNG to keep the alpha channel.\n\n## Parameters\n\n- prompt: what to generate. For RGBA, see Note: transparent.\n- negative_prompt: unused while cfg is 1.\n- cfg: keep 1 for the Qwen Image 2.1 official path. Raise it only if you use a negative prompt.\n- steps: Qwen Image 2.1 official pipeline uses about 40-50 with euler. This template starts at 25. More advanced samplers need fewer steps.\n" + }, + "color": "#222", + "bgcolor": "#000" + } + } + ], + "unconfirmed": [ + "虚拟节点、子图边界、开关与循环展开语义需独立核查", + "未有可信节点定义;单位、范围、环境和实际产物未验证" + ] + } +} diff --git a/docs/hel-892/sixth-evidence/05.json b/docs/hel-892/sixth-evidence/05.json new file mode 100644 index 00000000..d8ef2f93 --- /dev/null +++ b/docs/hel-892/sixth-evidence/05.json @@ -0,0 +1,2887 @@ +{ + "binding": { + "id": "05", + "source_path": "example_ComfyUI_workflows/4-视频生成/Minimax-h3/MiniMax_H3_导演台时间线素材_双采样潜空间放大_多分段拼接.json", + "source_sha": "ad7d38328e79df4fc9e76664148e45644fa6e4c8", + "export_file": "05.api.json", + "source_sha256": "172eb8af3aca70082941427520fdcca06568957bab3d39101c5e9d6641b45e4a", + "export_sha256": "95a01c1ea27a8279a6dc202820505099a94161d91e0789189b61208b09007af8", + "original_filename": "MiniMax_H3_导演台时间线素材_双采样潜空间放大_多分段拼接.json", + "provider": "白栖知", + "attachment_id": "01a187e2-3664-76ba-bacc-507b9e8c468a", + "received_at": "2026-10-29T13:38:38Z", + "received_at_basis": "主卡附件评论创建时间;不是导出时间", + "source_comment_id": "01a187e2-4f6e-761e-9412-43042e094daa", + "exported_at": null, + "frontend_version": null, + "node_versions": null, + "provenance": "user_supplied_api_export_environment_unknown", + "source_export_equivalence_verified": false, + "instruction": "视频时间线:7.timeline_data保留完整JSON字符串,7.finite_plan供8.finite_plan;8接模型/CLIP/双VAE/sampler/sigmas,9接图像及音频后送10保存。先核对时间线素材引用、generation_seconds、分辨率及seed;标题提及的二采与拼接不能代替实际16节点API证据,时间线执行语义待节点定义。" + }, + "adopted": true, + "method": "supplied_api_companion", + "general_conversion_verified": false, + "execution_verified": false, + "comparison": { + "equivalent": false, + "scope": "逐节点/字段保留API值和类型;直接连线比较ID及槽;位置控件不按顺序猜名", + "nodes": [ + { + "api_id": "1", + "api_node": { + "inputs": { + "unet_name": "MiniMax/minimax_h3_ref2va_pruned_int8_convrot.safetensors", + "weight_dtype": "default" + }, + "class_type": "UNETLoader", + "_meta": { + "title": "UNet加载器" + } + }, + "fields": [ + { + "field": "unet_name", + "api_value": "MiniMax/minimax_h3_ref2va_pruned_int8_convrot.safetensors", + "api_type": "str", + "named_evidence": { + "source_value": "MiniMax\\minimax_h3_ref2va_pruned_int8_convrot.safetensors", + "source_type": "str", + "status": "different" + }, + "source_value": "MiniMax\\minimax_h3_ref2va_pruned_int8_convrot.safetensors", + "status": "named_difference" + }, + { + "field": "weight_dtype", + "api_value": "default", + "api_type": "str", + "named_evidence": { + "source_value": "default", + "source_type": "str", + "status": "equal" + }, + "source_value": "default", + "status": "named_equal" + } + ], + "source_node": { + "id": 1, + "type": "UNETLoader", + "pos": [ + -430, + -980 + ], + "size": [ + 540, + 90 + ], + "flags": {}, + "order": 4, + "mode": 0, + "inputs": [ + { + "localized_name": "UNet名称", + "name": "unet_name", + "type": "COMBO", + "widget": { + "name": "unet_name" + }, + "link": null + }, + { + "localized_name": "数据类型", + "name": "weight_dtype", + "type": "COMBO", + "widget": { + "name": "weight_dtype" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "模型", + "name": "MODEL", + "type": "MODEL", + "links": [ + 36 + ] + } + ], + "properties": { + "Node name for S&R": "UNETLoader", + "cnr_id": "comfy-core", + "ver": "0.34.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "MiniMax\\minimax_h3_ref2va_pruned_int8_convrot.safetensors", + "default" + ], + "widgets_values_named": { + "unet_name": "MiniMax\\minimax_h3_ref2va_pruned_int8_convrot.safetensors", + "weight_dtype": "default" + } + }, + "status": "type_match", + "metadata": { + "source_title": "UNETLoader", + "api_meta": { + "title": "UNet加载器" + }, + "title_equal": false + } + }, + { + "api_id": "2", + "api_node": { + "inputs": { + "clip_name": "MiniMax/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors", + "type": "minimax", + "device": "default" + }, + "class_type": "CLIPLoader", + "_meta": { + "title": "加载CLIP" + } + }, + "fields": [ + { + "field": "clip_name", + "api_value": "MiniMax/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors", + "api_type": "str", + "named_evidence": { + "source_value": "MiniMax\\qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors", + "source_type": "str", + "status": "different" + }, + "source_value": "MiniMax\\qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors", + "status": "named_difference" + }, + { + "field": "type", + "api_value": "minimax", + "api_type": "str", + "named_evidence": { + "source_value": "minimax", + "source_type": "str", + "status": "equal" + }, + "source_value": "minimax", + "status": "named_equal" + }, + { + "field": "device", + "api_value": "default", + "api_type": "str", + "named_evidence": { + "source_value": "default", + "source_type": "str", + "status": "equal" + }, + "source_value": "default", + "status": "named_equal" + } + ], + "source_node": { + "id": 2, + "type": "CLIPLoader", + "pos": [ + -410, + -630 + ], + "size": [ + 530, + 110 + ], + "flags": {}, + "order": 0, + "mode": 0, + "inputs": [ + { + "localized_name": "CLIP名称", + "name": "clip_name", + "type": "COMBO", + "widget": { + "name": "clip_name" + }, + "link": null + }, + { + "localized_name": "类型", + "name": "type", + "type": "COMBO", + "widget": { + "name": "type" + }, + "link": null + }, + { + "localized_name": "设备", + "name": "device", + "shape": 7, + "type": "COMBO", + "widget": { + "name": "device" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "CLIP", + "name": "CLIP", + "type": "CLIP", + "links": [ + 14 + ] + } + ], + "properties": { + "Node name for S&R": "CLIPLoader", + "cnr_id": "comfy-core", + "ver": "0.34.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "MiniMax\\qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors", + "minimax", + "default" + ], + "widgets_values_named": { + "clip_name": "MiniMax\\qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors", + "type": "minimax", + "device": "default" + } + }, + "status": "type_match", + "metadata": { + "source_title": "CLIPLoader", + "api_meta": { + "title": "加载CLIP" + }, + "title_equal": false + } + }, + { + "api_id": "3", + "api_node": { + "inputs": { + "vae_name": "MiniMax/minimax_h3_video_vae_fp16.safetensors" + }, + "class_type": "VAELoader", + "_meta": { + "title": "加载VAE" + } + }, + "fields": [ + { + "field": "vae_name", + "api_value": "MiniMax/minimax_h3_video_vae_fp16.safetensors", + "api_type": "str", + "named_evidence": { + "source_value": "MiniMax\\minimax_h3_video_vae_fp16.safetensors", + "source_type": "str", + "status": "different" + }, + "source_value": "MiniMax\\minimax_h3_video_vae_fp16.safetensors", + "status": "named_difference" + } + ], + "source_node": { + "id": 3, + "type": "VAELoader", + "pos": [ + -400, + -420 + ], + "size": [ + 440, + 80 + ], + "flags": {}, + "order": 3, + "mode": 0, + "inputs": [ + { + "localized_name": "vae名称", + "name": "vae_name", + "type": "COMBO", + "widget": { + "name": "vae_name" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "VAE", + "name": "VAE", + "type": "VAE", + "links": [ + 46 + ] + } + ], + "properties": { + "Node name for S&R": "VAELoader", + "cnr_id": "comfy-core", + "ver": "0.34.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "MiniMax\\minimax_h3_video_vae_fp16.safetensors" + ], + "widgets_values_named": { + "vae_name": "MiniMax\\minimax_h3_video_vae_fp16.safetensors" + } + }, + "status": "type_match", + "metadata": { + "source_title": "VAELoader", + "api_meta": { + "title": "加载VAE" + }, + "title_equal": false + } + }, + { + "api_id": "4", + "api_node": { + "inputs": { + "vae_name": "MiniMax/minimax_h3_audio_vae_fp32.safetensors" + }, + "class_type": "VAELoader", + "_meta": { + "title": "加载VAE" + } + }, + "fields": [ + { + "field": "vae_name", + "api_value": "MiniMax/minimax_h3_audio_vae_fp32.safetensors", + "api_type": "str", + "named_evidence": { + "source_value": "MiniMax\\minimax_h3_audio_vae_fp32.safetensors", + "source_type": "str", + "status": "different" + }, + "source_value": "MiniMax\\minimax_h3_audio_vae_fp32.safetensors", + "status": "named_difference" + } + ], + "source_node": { + "id": 4, + "type": "VAELoader", + "pos": [ + -400, + -240 + ], + "size": [ + 440, + 60 + ], + "flags": {}, + "order": 1, + "mode": 0, + "inputs": [ + { + "localized_name": "vae名称", + "name": "vae_name", + "type": "COMBO", + "widget": { + "name": "vae_name" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "VAE", + "name": "VAE", + "type": "VAE", + "links": [ + 16 + ] + } + ], + "properties": { + "Node name for S&R": "VAELoader", + "cnr_id": "comfy-core", + "ver": "0.34.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "MiniMax\\minimax_h3_audio_vae_fp32.safetensors" + ], + "widgets_values_named": { + "vae_name": "MiniMax\\minimax_h3_audio_vae_fp32.safetensors" + } + }, + "status": "type_match", + "metadata": { + "source_title": "VAELoader", + "api_meta": { + "title": "加载VAE" + }, + "title_equal": false + } + }, + { + "api_id": "5", + "api_node": { + "inputs": { + "scheduler": "simple", + "steps": 8, + "denoise": 1, + "model": [ + "22", + 0 + ] + }, + "class_type": "BasicScheduler", + "_meta": { + "title": "基本调度器" + } + }, + "fields": [ + { + "field": "scheduler", + "api_value": "simple", + "api_type": "str", + "named_evidence": { + "source_value": "simple", + "source_type": "str", + "status": "equal" + }, + "source_value": "simple", + "status": "named_equal" + }, + { + "field": "steps", + "api_value": 8, + "api_type": "int", + "named_evidence": { + "source_value": 8, + "source_type": "int", + "status": "equal" + }, + "source_value": 8, + "status": "named_equal" + }, + { + "field": "denoise", + "api_value": 1, + "api_type": "int", + "named_evidence": { + "source_value": 1, + "source_type": "int", + "status": "equal" + }, + "source_value": 1, + "status": "named_equal" + }, + { + "field": "model", + "api_value": [ + "22", + 0 + ], + "api_type": "list", + "source_link": [ + 41, + 22, + 0, + 5, + 0, + "MODEL" + ], + "source_expected": [ + "22", + 0 + ], + "status": "link_equal" + } + ], + "source_node": { + "id": 5, + "type": "BasicScheduler", + "pos": [ + 674.924866337904, + -488.18032078903093 + ], + "size": [ + 270, + 110 + ], + "flags": {}, + "order": 13, + "mode": 0, + "inputs": [ + { + "localized_name": "模型", + "name": "model", + "type": "MODEL", + "link": 41 + }, + { + "localized_name": "调度器", + "name": "scheduler", + "type": "COMBO", + "widget": { + "name": "scheduler" + }, + "link": null + }, + { + "localized_name": "步数", + "name": "steps", + "type": "INT", + "widget": { + "name": "steps" + }, + "link": null + }, + { + "localized_name": "降噪", + "name": "denoise", + "type": "FLOAT", + "widget": { + "name": "denoise" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "Sigmas", + "name": "SIGMAS", + "type": "SIGMAS", + "links": [ + 19 + ] + } + ], + "properties": { + "Node name for S&R": "BasicScheduler", + "cnr_id": "comfy-core", + "ver": "0.34.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "simple", + 8, + 1 + ], + "widgets_values_named": { + "scheduler": "simple", + "steps": 8, + "denoise": 1 + } + }, + "status": "type_match", + "metadata": { + "source_title": "BasicScheduler", + "api_meta": { + "title": "基本调度器" + }, + "title_equal": false + } + }, + { + "api_id": "6", + "api_node": { + "inputs": { + "sampler_name": "euler" + }, + "class_type": "KSamplerSelect", + "_meta": { + "title": "K采样器选择" + } + }, + "fields": [ + { + "field": "sampler_name", + "api_value": "euler", + "api_type": "str", + "named_evidence": { + "source_value": "euler", + "source_type": "str", + "status": "equal" + }, + "source_value": "euler", + "status": "named_equal" + } + ], + "source_node": { + "id": 6, + "type": "KSamplerSelect", + "pos": [ + 682.3753995974032, + -603.9644214539213 + ], + "size": [ + 252.3294296811564, + 58 + ], + "flags": {}, + "order": 5, + "mode": 0, + "inputs": [ + { + "localized_name": "采样器名称", + "name": "sampler_name", + "type": "COMBO", + "widget": { + "name": "sampler_name" + }, + "link": null + } + ], + "outputs": [ + { + "localized_name": "采样器", + "name": "SAMPLER", + "type": "SAMPLER", + "links": [ + 18 + ] + } + ], + "properties": { + "Node name for S&R": "KSamplerSelect", + "cnr_id": "comfy-core", + "ver": "0.34.0", + "ue_properties": { + "widget_ue_connectable": {}, + "version": "7.8", + "input_ue_unconnectable": {} + } + }, + "widgets_values": [ + "euler" + ], + "widgets_values_named": { + "sampler_name": "euler" + } + }, + "status": "type_match", + "metadata": { + "source_title": "KSamplerSelect", + "api_meta": { + "title": "K采样器选择" + }, + "title_equal": false + } + }, + { + "api_id": "7", + "api_node": { + "inputs": { + "width": [ + "24", + 0 + ], + "height": [ + "24", + 1 + ], + "generation_seconds": 11.5, + "timeline_data": "{\"version\":9,\"fps\":24,\"globalPrompt\":\"\",\"secondPass\":true,\"secondPassModel\":\"minimax_h3_latent_upscaler_3d_conv_v1_fp16.safetensors\",\"secondPassHighSteps\":2,\"selection\":{\"start\":8.5,\"duration\":11.541666666666666},\"videoAudioEnabled\":true,\"videoClips\":[],\"images\":[{\"id\":\"mu5mo29c-dme5zh\",\"file\":\"minimax_h3_timeline_director/1789655485241_lz840_20260324_215319_Z-Image_00001_.png\",\"name\":\"20260324_215319_Z-Image_00001_.png\",\"width\":1080,\"height\":1920},{\"id\":\"mu5moaxk-7a9fy1\",\"file\":\"minimax_h3_timeline_director/1789655496464_mnjqz_20260324_214939_Z-Image_00001_.png\",\"name\":\"20260324_214939_Z-Image_00001_.png\",\"width\":1080,\"height\":1920}],\"audios\":[],\"segmentConfig\":{\"count\":2,\"activeIndex\":1,\"mode\":\"timeline\",\"segments\":[{\"images\":[\"mu5mo29c-dme5zh\"],\"audios\":[],\"prompt\":\"subject_definitions:\\n is the yellow-furred bear in, with light golden-yellow fur, a brown military-style canvas jacket with brass buttons and worn shoulder patches, round blue eyes, a large dark-brown nose, and a tense furious expression.\\n is the red-furred bear in, with deep russet-red fur, a red traditional Chinese-collar tunic with knotted frog closures down the front, round green eyes, a large dark-red nose, and a cold rage expression.\\n