Fetches wiki pages from all core/* repos on Forge, parses into
sections, and stores as type:service with repo/lang tags. Gives
the PHP orchestrator contextual knowledge about the Go services
it coordinates.
71+ pages across 22+ repos, ~770 vectorised sections.
Co-Authored-By: Virgil <virgil@lethean.io>
Discovers markdown files from core-php/docs/build/php/ and packages/
for vectorisation into OpenBrain. Tagged as source:docs with 0.85
confidence, typed as documentation.
Co-Authored-By: Virgil <virgil@lethean.io>
- Truncate content to 3800 chars before embedding (embeddinggemma has
a 2048-token context, ~4K char limit). Eliminates all 73 Ollama 500
errors from oversized plan sections.
- Clear brain_memories DB table when --fresh is used, keeping DB rows
in sync with Qdrant vectors.
Co-Authored-By: Virgil <virgil@lethean.io>
Discovers markdown across 4 source types:
- memory: ~/.claude/projects memory files
- plans: docs/plans across repos + ~/.claude/plans
- claude-md: CLAUDE.md repo instructions
- tasks: core/tasks research and ideas
Supports --fresh to clear collection, --dry-run for preview,
and --source to target specific types.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
