本文へスキップ
Kitbay
スキル公式

mongodb-natural-language-querying

Apache-2.02026年11月18日 更新206GitHubスター

MongoDB's skill for turning a plain-language request into a read-only MongoDB find query or aggregation pipeline.

スキルの内容

MongoDB's skill for turning a plain-language request into a read-only MongoDB find query or aggregation pipeline. Through the MongoDB MCP Server the agent first collects the collection's indexes, schema and a few sample documents, then writes a query that fits them, which can include geospatial queries, text matching and lookups across collections. It also helps convert SQL-style requests. It leaves out Atlas Search and vector search, analysis of existing slow queries, and any pipeline that writes data, pointing to other MongoDB skills for those. It requires the MCP server.

できること

  • Write find queries from a plain-language description
  • Build aggregation pipelines using the collection's schema and indexes
  • Translate SQL-style requests into MongoDB queries
  • Inspect sample documents and indexes before writing a query