| name | zeabur-rag-search |
| description | Search the RAG knowledge base via hybrid search (semantic + BM25). Use whenever answering technical questions that may be covered in your knowledge base — even if the user doesn't explicitly ask you to search. Also use when you need to verify claims or find up-to-date information. |
RAG Search
Search your knowledge base using hybrid retrieval (semantic + BM25 keyword).
Base URL: $ZEABUR_RAG_URL
Auth: Authorization: Bearer $RAG_API_KEY
API
curl -s -X POST "$ZEABUR_RAG_URL/api/query" \
-H "Authorization: Bearer $RAG_API_KEY" \
-H "Content-Type: application/json" \
-d '{"query": "your question", "mode": "hybrid", "rag": false, "top_k": 5, "client": "claude-code"}'
| Parameter | Default | Description |
|---|
query | required | Search query (supports multilingual) |
mode | "hybrid" | "hybrid", "semantic", or keyword-only "bm25" |
top_k | 5 | Number of chunks to return |
rag | false | If true, also returns LLM-generated answer string |
keyword_weight | 0.25 | BM25 weight in hybrid mode |
semantic_weight | 0.75 | Semantic weight in hybrid mode |
rerank | false | If true, use an LLM to reorder a wider candidate pool |
rerank_pool_size | 30 | Candidate count considered when reranking |
decay | 180 | Temporal decay half-life in days (0 = off) |
stream | false | Stream the RAG answer (NDJSON) |
sources | key permissions | Optional source list, intersected with the API key's read permissions |
client | "api" | Caller identifier for analytics (e.g. "claude-code", "web-ui", "mcp") |
Always pass client: "claude-code" when calling from Claude Code so queries are trackable in the dashboard.
Response includes chunks[] (each with id, title, answer, tags, similarity (relevance score), source, verified) and signal_id for linking feedback. Score scales differ between BM25, hybrid RRF, and semantic modes, so do not compare values across modes.
Tips
- Use
rag: false when you only need raw chunks to read yourself — it's faster and cheaper.
- Set
rag: true only when you want the service to generate a summarized answer.
- The
signal_id in the response can be passed to the feedback API to rate result quality.
