Task
\core/rag.py\ currently ingests plain strings via \ingest_document. Issue #22 asked for PDF/txt/markdown file ingestion; the RAG plugin (\plugins/builtins/rag_plugin.py) is text-only today.
Do
- Extend \ingest_document\ to accept a \source\ type, or add a helper to read local files
- Add markdown handling (strip code fences/headings as needed) and .txt\ support
- Wire it into the \ingest_document\ tool / API so users can pass a file path
- Add tests in \ ests/test_rag.py\ (mirror the existing tmp_path-isolated pattern)
Verify
- \python -m pytest tests/test_rag.py -q\ passes
- Ingesting a small .md\ file returns sensible chunks
uff check .\ clean
Notes
- PDF is harder (needs an extra lib) — markdown + txt is the core scope; note PDF as future work.
- Use \�syncio.to_thread\ for file reads per project conventions.
Task
\core/rag.py\ currently ingests plain strings via \ingest_document. Issue #22 asked for PDF/txt/markdown file ingestion; the RAG plugin (\plugins/builtins/rag_plugin.py) is text-only today.
Do
Verify
uff check .\ clean
Notes