hybrid-search-rrf
Reciprocal Rank Fusion for merging full-text and vector search results in this project. Convex native searchIndex + vectorIndex integration with admin-toggleable hybrid search. Triggers on "hybrid search", "RRF", "full-text", "vector search", "mergeMessagesWithRRF", "searchIndex", "vectorIndex".
npx skills add majiayu000/claude-skill-registry --skill hybrid-search-rrf --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
# Hybrid Search with RRF Project combines Convex full-text search (keyword matching) with vector search (semantic similarity) using Reciprocal Rank Fusion algorithm. Admin-controlled feature for messages and memories. ## Why Hybrid Search **Coverage gap**: Full-text finds exact keywords, vector finds meaning. Neither is complete. Examples: - User searches "cancel subscription" - vector finds "terminate my plan" - User searches "GPT-4" - full-text finds exact model name missed by embeddings - Overlapping results get boosted (appear in both = high relevance) **When enabled**: Check admin settings in actions, fallback to text-only if disabled. ```typescript // From convex/search/hybrid.ts const adminSettings = (await (ctx.runQuery as any)( api.adminSettings.get, {}, )) as { enableHybridSearch?: boolean } | null; const hybridSearchEnabled = adminSettings?.enableHybridSearch ?? false; if (!hybridSearchEnabled) { return textResults.slice(0, limit); } ``` ## RRF Algorithm **Formula**: `score = weight * 1 / (k + rank + 1)` - `k = 60` (default constant, smooths ranking differences) - Higher rank = lower score - Overlapping results: sum scores from both sources (boosted) - Optional source we
- Why Hybrid Search
- RRF Algorithm
- Full-Text Search (Convex searchIndex)
- Vector Search (Convex vectorIndex)
- Hybrid Search Action (Messages)
- Hybrid Search with Reranking (Memories)
- Fallback Pattern
- Quality Thresholds
- Key Files
- Common Patterns
- Avoid
What does the hybrid-search-rrf skill do?
Reciprocal Rank Fusion for merging full-text and vector search results in this project. Convex native searchIndex + vectorIndex integration with admin-toggleable hybrid search. Triggers on "hybrid search", "RRF", "full-text", "vector search", "mergeMessagesWithRRF", "searchIndex", "vectorIndex".
How do I install it?
Run `npx skills add majiayu000/claude-skill-registry --skill hybrid-search-rrf --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.