fix(skill): stamp _origin on the semantic tier so update cannot delete it (#2843)

graphify/cli.py

@@ -3898,6 +3898,14 @@ def _progress(idx: int, total: int, _result: dict) -> None:
         # first means semantic node attributes win on collision (richer labels
         # for symbols also referenced in docs). Hyperedges only come from the
         # semantic side.
+        # Everything in sem_result is semantic tier - fresh results and cache
+        # hits alike - so say so before the merge (#2843): an unstamped item
+        # is classified by the shape of its source_location, and a doc node
+        # located at 'L12' would read as AST and be deleted by the next
+        # incremental rebuild. Fresh results already carry the stamp
+        # (llm._stamp_semantic_origin); cache entries from before it do not.
+        from graphify.llm import _stamp_semantic_origin as _stamp_sem
+        _stamp_sem(sem_result)
         merged: dict = {
             "nodes": list(ast_result.get("nodes", [])) + list(sem_result.get("nodes", [])) + list(pg_result.get("nodes", [])) + list(cargo_result.get("nodes", [])),
             "edges": list(ast_result.get("edges", [])) + list(sem_result.get("edges", [])) + list(pg_result.get("edges", [])) + list(cargo_result.get("edges", [])),

graphify/llm.py

@@ -2794,6 +2794,7 @@ def _out_of_scope(item: dict) -> bool:
         if p.resolve() not in {c.resolve() for c in covered}
     )
     merged["uncovered_files"] = [str(p) for p in uncovered]
+    _stamp_semantic_origin(merged)
     if uncovered:
         shown = ", ".join(p.name for p in uncovered[:5])
         more = f" (+{len(uncovered) - 5} more)" if len(uncovered) > 5 else ""
@@ -2806,6 +2807,24 @@ def _out_of_scope(item: dict) -> bool:
     return merged
 
 
+def _stamp_semantic_origin(result: dict) -> dict:
+    """Mark every node and edge of a semantic extraction ``_origin: "semantic"``.
+
+    build._is_ast_tier reads ``_origin`` when present and otherwise guesses
+    from the shape of ``source_location``: ``L<line>`` means AST. A backend
+    or subagent that reports line numbers for a document's sections made
+    those nodes read as AST, and the next incremental rebuild deleted them
+    with the re-extracted code file they seemed to belong to (#2843).
+    extract() stamps its items ``ast``; this is the semantic counterpart.
+    Existing stamps are kept (``setdefault``); returns ``result``.
+    """
+    for key in ("nodes", "edges"):
+        for item in result.get(key, []) or []:
+            if isinstance(item, dict):
+                item.setdefault("_origin", "semantic")
+    return result
+
+
 def _merge_into(merged: dict, result: dict) -> None:
     """Append a chunk result into the running merged accumulator."""
     merged["nodes"].extend(result.get("nodes", []))

graphify/skill-agents.md

@@ -363,6 +363,18 @@ from pathlib import Path
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text(encoding=\"utf-8\"))
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text(encoding=\"utf-8\"))
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill-aider.md

@@ -361,6 +361,18 @@ from pathlib import Path
 ast = json.loads(Path('.graphify_ast.json').read_text())
 sem = json.loads(Path('.graphify_semantic.json').read_text())
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill-amp.md

@@ -363,6 +363,18 @@ from pathlib import Path
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text(encoding=\"utf-8\"))
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text(encoding=\"utf-8\"))
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill-claw.md

@@ -366,6 +366,18 @@ from pathlib import Path
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text(encoding=\"utf-8\"))
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text(encoding=\"utf-8\"))
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill-codex.md

@@ -363,6 +363,18 @@ from pathlib import Path
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text(encoding=\"utf-8\"))
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text(encoding=\"utf-8\"))
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill-copilot.md

@@ -366,6 +366,18 @@ from pathlib import Path
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text(encoding=\"utf-8\"))
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text(encoding=\"utf-8\"))
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill-devin.md

@@ -425,6 +425,18 @@ from graphify.semantic_cleanup import sanitize_semantic_fragment
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text())
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text())
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill-droid.md

@@ -363,6 +363,18 @@ from pathlib import Path
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text(encoding=\"utf-8\"))
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text(encoding=\"utf-8\"))
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill-kilo.md

@@ -366,6 +366,18 @@ from pathlib import Path
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text(encoding=\"utf-8\"))
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text(encoding=\"utf-8\"))
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill-kiro.md

@@ -366,6 +366,18 @@ from pathlib import Path
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text(encoding=\"utf-8\"))
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text(encoding=\"utf-8\"))
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill-opencode.md

@@ -358,6 +358,18 @@ from pathlib import Path
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text(encoding=\"utf-8\"))
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text(encoding=\"utf-8\"))
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill-pi.md

@@ -366,6 +366,18 @@ from pathlib import Path
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text(encoding=\"utf-8\"))
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text(encoding=\"utf-8\"))
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill-trae.md

@@ -364,6 +364,18 @@ from pathlib import Path
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text(encoding=\"utf-8\"))
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text(encoding=\"utf-8\"))
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill-vscode.md

@@ -362,6 +362,18 @@ from pathlib import Path
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text(encoding=\"utf-8\"))
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text(encoding=\"utf-8\"))
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill-windows.md

@@ -393,6 +393,18 @@ from pathlib import Path
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text(encoding="utf-8"))
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text(encoding="utf-8"))
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])

graphify/skill.md

@@ -366,6 +366,18 @@ from pathlib import Path
 ast = json.loads(Path('graphify-out/.graphify_ast.json').read_text(encoding=\"utf-8\"))
 sem = json.loads(Path('graphify-out/.graphify_semantic.json').read_text(encoding=\"utf-8\"))
 
+# Stamp tier provenance on the semantic side. build._is_ast_tier reads
+# _origin when present and otherwise guesses from the SHAPE of
+# source_location ('L<line>' means AST). A subagent that writes 'L12' for a
+# doc section therefore made that node read as AST, and the next
+# `graphify update` deleted it along with the re-extracted code file it
+# seemed to belong to (#2843). extract() stamps its own items 'ast'; the
+# semantic side must say so explicitly.
+for n in sem['nodes']:
+    n.setdefault('_origin', 'semantic')
+for e in sem['edges']:
+    e.setdefault('_origin', 'semantic')
+
 # Merge: AST nodes first, semantic nodes deduplicated by id
 seen = {n['id'] for n in ast['nodes']}
 merged_nodes = list(ast['nodes'])