AnyDoc Office & Document Curator Skill

When this skill is invoked via /okc-doc <input_path>, execute the following 2-step streamlined workflow:

1. Extract & Stage Document Content (FastMCP — 1 Turn):
   • Execute the native FastMCP tool fetch_doc_data(file_path="<input_path>", slug="<slug>") on the okc server (do not spawn CLI commands unless MCP is unavailable).
   • Features: Multi-format parsing (.docx, .pptx, .xlsx, .epub, .pdf, .csv, .odt, .rtf, .txt), concurrent extraction of embedded diagrams to <VAULT_ROOT>/assets/images/, inline text extraction (up to MAX_TRANSCRIPT_CHARS), and SQLite graphify_context (author_existing_notes, matched_wiki_notes).
   • Only read temp/fetched_data.txt if is_truncated == True.
2. Synthesize & Atomically Commit (FastMCP — 1 Turn):
   • Consume the inline text and graphify_context returned by fetch_doc_data.
   • Format a curated note in the raw/ zone following vault conventions (no YAML, H1 title, blockquote with Processed: DD-MM-YYYY and Tags: #no-read-yet, Key Takeaways, numbered sections, flashcards, glossary, related wikilinks).
   • Identify 3-7 core concepts and format their markdown content for wiki/ (evolving inline matched_wiki_notes where applicable).
   • Commit all notes, surgically upsert into SQLite, and trigger the detached background worker (async_post_commit.py) in a single call using FastMCP commit_curated_note:

     commit_curated_note(
         raw_note_path="dataScienceKnowledgeBase/AI Engineer/raw/<Category>/<Document Title>.md",
         raw_content="...",
         wiki_notes={"ConceptName.md": "..."},
         target_kb="dataScienceKnowledgeBase/AI Engineer",
         fast_sync=True,
         async_sync=True
     )

   • Zero-Spam Policy: Perform any vault checks silently using native cortex tools (find_by_name, view_file, find_notes). Never run exploratory terminal commands (python3 -c, ls, etc.).
   • Report progress and show the Obsidian URI to the user.