NCBI Entrez MCP Server

{ "name": "staged-sql-analysis", "description": "Guide for staging Entrez records and exploring them with SQL through the durable object pipeline.", "instructions": [ "Call `entrez_capabilities` (alias `entrez-capabilities`) with { \"tool\": \"entrez_data\", \"format\": \"detailed\" } to review staging operations and required parameters.", "Stage source data via `entrez_data` with `operation: \"fetch_and_stage\"`. Provide `database`, `ids`, and optionally `rettype`. Capture the returned `data_access_id`.", "If you need iterative queries, list datasets using `operation: \"list_datasets\"` to confirm availability and retention policy.", "Run SQL using `operation: \"query\"` with either a user-supplied `sql` statement or `smart_summary: true` plus `intended_use` hints when you want auto-generated insights.", "Inspect schema details using `operation: \"schema\"` before crafting complex queries." ], "retry_guidance": [ "Keep the staged payload modest—paginate fetch operations or limit `ids` when the dataset becomes large.", "If SQL fails, check the schema output for exact column names and types.", "Use `smart_summary: true` when you are unsure which SQL to run first." ], "primary_tools": ["entrez_data", "entrez_query", "entrez_capabilities"] }