Patrick Plate
155d56e8e8
- Move bigmind/ -> mcp/bigmind/ - Move webscraper/ -> mcp/webscraper/ - Move mss-failsafe/ -> java/mss-failsafe/ - Move Wellmann-Shop/ -> java/wellmann-shop/ (normalize to kebab-case) - Add .roo/ IDE config files to tracking - Add plans/REPO_STRATEGY.md (monorepo strategy document) - Expand .gitignore: Java/Maven, Node/TS, coverage, uv.lock - Rewrite README.md as navigation index - Update .roo/mcp.json webscraper path to mcp/webscraper/
1.2 KiB
1.2 KiB
BigMind Search Optimization — FTS5 Rules
FTS5 AND-Match Behavior
- Every token in query must appear in the same fact/chunk
- Order doesn't matter, but all words must match
- Case-insensitive
Good Query Patterns (2-3 Focused Keywords)
| ✅ Good | ❌ Bad | Why Good |
|---|---|---|
"TrueNAS Docker" |
"homelab infrastructure TrueNAS Docker" |
Too many tokens → 0 results |
"mcp.json config" |
"mcp.json clients VS Code IntelliJ config path" |
Rare words like "clients" kill results |
"Fedora workstation" |
"server infrastructure" |
Specific nouns > generic words |
"BigMind hypothesis" |
"memory_add_hypothesis during analysis" |
Focus on tool name + context |
Reserved Word Protection
- FTS5 keywords (rank, content, category) are auto-quoted
- Always quote multi-word queries:
"multi word query"
Search Strategy
- Start with 2 keywords, add third if needed
- Use tool names + context:
"memory_search_facts facts" - For people:
"memory_recall_person Patrick" - For sessions:
"memory_list_sessions mcp-adp"
When Search Returns 0
- Shorten query to 1-2 tokens
- Use
memory_list_sessions(topics_filter="mcp")for broad exploration - Fall back to
memory_get_context()for recent sessions