API

April 23, 2026 · View on GitHub

All methods are exposed as MCP tools.

search_docs

Unified search tool for documentation using semantic (vector similarity), keyword (BM25), or hybrid (both, merged with reciprocal rank fusion). Mode is selected with semanticWeight.

MCP Tool: search_docs

Input

{
  // required — corpus and optional version in one enum value:
  //   "tiger" | "postgres_14" … "postgres_18" | "postgis_3.3" … "postgis_3.6"
  "source": "postgres_17",
  "query": "How do I create an index?", // required
  "limit": 20, // optional; default 20 if omitted or null
  "semanticWeight": 0.7 // optional; 0–1 in 0.1 steps; default 0.7 if omitted or null
}
  • source: tiger (Tiger Cloud / TimescaleDB), postgres_XX for a specific PostgreSQL manual version, or postgis_X.X for PostGIS. There is no separate version field.
  • semanticWeight: 0 = keyword only, 1 = semantic only. Values strictly between 0 and 1 run semantic and keyword search in parallel (two DB queries), fuse ranked lists with RRF, and return rrf_score per row.
{
  "results": [
    {
      "id": 11716,
      "content": "CREATE INDEX ...",
      "metadata": "{...}", // JSON-encoded metadata
      "distance": 0.407 // lower = more relevant
    }
  ]
}
{
  "results": [
    {
      "id": 11716,
      "content": "CREATE INDEX ...",
      "metadata": "{...}", // JSON-encoded metadata
      "score": 12.5 // higher = more relevant
    }
  ]
}
{
  "results": [
    {
      "id": 11716,
      "content": "CREATE INDEX ...",
      "metadata": "{...}", // JSON-encoded metadata
      "rrf_score": 0.0328 // higher = more relevant; RRF of semantic + keyword ranks
    }
  ]
}

Skills

view_skill

Retrieves curated skills for common PostgreSQL and TimescaleDB tasks. This tool is disabled when deploying as a claude plugin (which use agent skills directly).

MCP Tool: view_skill

Input

{
  "name": "setup-timescaledb-hypertables", // see available skills in tool description
  "path": "SKILL.md", // optional, defaults to "SKILL.md"
}

Output

{
  "name": "setup-timescaledb-hypertables",
  "path": "SKILL.md",
  "description": "Step-by-step instructions for designing table schemas and setting up TimescaleDB with hypertables, indexes, compression, retention policies, and continuous aggregates.",
  "content": "...", // full skill content
}

Available Skills: Check the MCP tool description for the current list of available skills or look in the skills directory.