Loading tool details…
The Amazon Bedrock Knowledge Base Retrieval MCP Server is a free, Apache 2.0-licensed tool from AWS Labs that plugs Amazon Bedrock Knowledge Bases directly into any Model Context Protocol client — Claude Desktop, Cursor, VS Code (with the MCP extension), Windsurf, Cline, and AWS's own Kiro IDE. Instead of building custom retrieval-augmented-generation (RAG) plumbing, developers install a single package (`uvx awslabs.bedrock-kb-retrieval-mcp-server@latest`) and give their AI assistant natural-lan
The Amazon Bedrock Knowledge Base Retrieval MCP Server is a free, Apache 2.0-licensed tool from AWS Labs that plugs Amazon Bedrock Knowledge Bases directly into any Model Context Protocol client — Claude Desktop, Cursor, VS Code (with the MCP extension), Windsurf, Cline, and AWS's own Kiro IDE. Instead of building custom retrieval-augmented-generation (RAG) plumbing, developers install a single package (`uvx awslabs.bedrock-kb-retrieval-mcp-server@latest`) and give their AI assistant natural-lan
The Amazon Bedrock Knowledge Base Retrieval MCP Server is an open-source implementation of the Model Context Protocol (MCP) maintained by AWS Labs. It bridges AI assistants and coding agents (such as Claude Desktop, Cursor, Cline, and Amazon Q Developer) with Amazon Bedrock Knowledge Bases, enabling them to perform retrieval-augmented generation (RAG) against enterprise data stored in AWS without writing custom integration code.
At its core, the server exposes a small set of MCP tools that let an AI client discover available knowledge bases, list their underlying data sources (S3 buckets, Confluence, SharePoint, Salesforce, web crawlers, etc.), and run natural-language queries against them. Results are returned with citations pointing back to the original source documents, which is essential for verifiable answers in regulated and enterprise settings. The server supports filtering by data source so that a single knowledge base containing multiple repositories can be queried selectively, and it integrates with Amazon Bedrock's reranking models to push the most relevant chunks to the top of the result set before they reach the LLM.
Because the server runs locally (typically launched via uvx or a similar Python runner) and authenticates using standard AWS credentials, all access to knowledge bases is governed by IAM policies. This means existing access controls, audit logging via CloudTrail, and VPC networking constraints continue to apply — the MCP server does not introduce a new authorization layer. Knowledge bases must be tagged with mcp-multirag-kb=true to be discoverable, giving administrators a simple opt-in mechanism for which corpora are exposed to AI agents.
The server is designed for developers building agentic applications, internal copilots, customer support assistants, and documentation chatbots that need grounded, citation-backed responses from proprietary data. It removes a significant amount of boilerplate around the Bedrock Agent Runtime APIs (Retrieve and RetrieveAndGenerate) and standardizes the contract so that any MCP-compatible client can consume the same knowledge base with no client-specific code. As part of the broader awslabs/mcp project on GitHub, it is released under the Apache 2.0 license and ships alongside companion servers for Bedrock AgentCore, SageMaker, Amazon Q Index, and other AWS AI services, making it a natural building block for teams already standardized on AWS.
Free (Apache 2.0)
Pay-per-query
Ready to get started with Amazon Bedrock Knowledge Base Retrieval MCP Server?
View Pricing Options →We believe in transparent reviews. Here's what Amazon Bedrock Knowledge Base Retrieval MCP Server doesn't handle well:
In 2026, the awslabs/mcp project has expanded around this server with companion servers for Bedrock AgentCore, SageMaker AI, Amazon Q Index, and a dedicated Document Loader MCP Server, letting teams pair retrieval with ingestion and agent orchestration entirely through MCP. The Bedrock KB server itself has tracked recent Bedrock Agent Runtime updates including expanded reranking model support, finer-grained metadata filters, and improved handling of multi-source knowledge bases. The broader AWS Labs MCP catalog now ships with installation guidance for Cursor, Cline, Claude Desktop, and Amazon Q Developer, reflecting MCP's rapid adoption as the default integration surface for enterprise AI assistants.
Pricing and features change often. If anything on this Amazon Bedrock Knowledge Base Retrieval MCP Server page is wrong or missing, tell us and include a link to the source. We review every correction.
I ran a six-week evaluation of 15 AI developer tools across four workloads: a Flask-to-FastAPI refactor, a Next.js dashboard built from scratch, 23 bug-fix issues sampled from SWE-bench Verified, and a Rust-to-Go port of a 600-line CLI. **Rankings reflect what shipped working cod
Hidden gems in the AI agent tooling space — from browser infrastructure to memory platforms to observability tools. These production-ready tools solve real problems that most developers haven't discovered yet.