io.github.HatmanStack/ragstacksearch

ragstack

Verified · yesterday

Search, chat, upload, and scrape a serverless RAGStack knowledge base on AWS.

Install

claude mcp add ragstack -- uvx ragstack-mcp

Requires: RAGSTACK_GRAPHQL_ENDPOINT, RAGSTACK_API_KEY

Our take

The MCP companion to RAGStack-Lambda, a scale-to-zero serverless RAG stack on AWS (Lambda, Bedrock, S3, Transcribe) that you deploy yourself. Once your stack is up, the server gives agents search, chat, upload, and scrape against your own knowledge base with no idle costs — a real alternative to paying for a hosted vector DB. The caveat is the prerequisite: this is a thin GraphQL client, and the value lives in deploying and running the whole AWS stack first.

reviewed by hand · 2026-07-05

Something wrong or dead here? Report it

Verification record

last verified: yesterday
github stars: 25
last commit: 5 days ago
archived: no
license: Apache-2.0
in registry since: 2026-02-06

github.com/HatmanStack/RAGStack-Lambda