🔧#640 — LinkResolver selects worst match instead of best
Replace `min(results, key=lambda x: x.score)` with `results[0]`.
Both SQLite and Postgres return results sorted best-first in SQL,
so using `results[0]` is backend-agnostic and correct.
🔧#641 — search_notes output_format="text" returns raw Pydantic model
Add `_format_search_markdown()` that formats SearchResponse as readable
markdown with title, permalink, score, and matched snippet per result.
Update prompts to use `output_format="json"` since they need structured
data for result counting and branching logic.
🔧#642 — metadata_filters with `note_type` key returns empty results
Add `_METADATA_KEY_ALIASES` mapping at the tool level that aliases
`note_type` → `type` before passing metadata_filters to the search query.
The frontmatter field is `type`, not `note_type`.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: phernandez <paul@basicmachines.co>
- Prompts (search, continue_conversation) now call MCP tools directly
instead of going through API endpoints, matching the recent_activity
pattern and fixing #526 where prompts returned empty results
- sync_file catches SemanticDependenciesMissingError separately so
entities are still returned successfully when vector embedding fails,
with a clear warning instead of silent failure (#578)
- `bm status` and `bm doctor` default to local routing since they scan
the local filesystem — cloud routing returned Docker-internal paths
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: phernandez <paul@basicmachines.co>
This change relocates the AI assistant guide from the static directory
into the package resources directory, ensuring it gets properly included
in the distribution package and is accessible when installed via pip/uv.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: phernandez <paul@basicmachines.co>
feat: Add multiple projects support
feat: enhanced read_note for when initial result is not found
fix: merge frontmatter when updating note
fix: handle directory removed on sync watch
## Summary
- Add comprehensive documentation to all MCP prompt modules
- Enhance search prompt with detailed contextual output formatting
- Implement consistent logging and docstring patterns across prompt
utilities
- Fix type checking in prompt modules
## Prompts Added/Enhanced
- `search.py`: New formatted output with relevance scores, excerpts, and
next steps
- `recent_activity.py`: Enhanced with better metadata handling and
documentation
- `continue_conversation.py`: Improved context management
## Resources Added/Enhanced
- `ai_assistant_guide`: Resource with description to give to LLM to
understand how to use the tools
## Technical improvements
- Added detailed docstrings to all prompt modules explaining their
purpose and usage
- Enhanced the search prompt with rich contextual output that helps LLMs
understand results
- Created a consistent pattern for formatting output across prompts
- Improved error handling in metadata extraction
- Standardized import organization and naming conventions
- Fixed various type checking issues across the codebase
This PR is part of our ongoing effort to improve the MCP's interaction
quality with LLMs, making the system more helpful and intuitive for AI
assistants to navigate knowledge bases.
🤖 Generated with [Claude Code](https://claude.ai/code)
---------
Co-authored-by: phernandez <phernandez@basicmachines.co>
Paul Hernandez