Suggest a server AI tools directory →

AWS Knowledge Base MCP server

Last verified: 2026-09

Query an Amazon Bedrock knowledge base.

Official reference Cloud & infra Needs a secret Claude Desktop, Cursor, Windsurf, Claude Code, VS Code

What it does

Queries an Amazon Bedrock knowledge base and returns chunks. Read-only on the KB you configure. You still need AWS credentials that can call Bedrock.

It sits in Cloud & infra: Most of these are admin APIs.

Good for

  • Query an Amazon Bedrock knowledge base you already populated.
  • Retrieve internal docs the KB was built from — not create VPCs.
  • Keep the IAM role to that KB only.
  • Skip it if the docs are already in Notion or Drive.

Tools

  • retrieve_from_aws_kb — Semantic retrieve over the knowledge base.

Config (Claude Desktop / Cursor / Windsurf)

Paste the JSON below. Same mcpServers shape. Replace placeholder paths and secrets. Windows paths look like C:\\Users\\you\\project, not /path/to. Host-side steps →

Claude Desktop

OSConfig file
macOS~/Library/Application Support/Claude/claude_desktop_config.json
Windows%APPDATA%\Claude\claude_desktop_config.json (usually C:\Users\<you>\AppData\Roaming\Claude\)
Linux~/.config/Claude/claude_desktop_config.json

Cursor

ScopemacOS / LinuxWindows
This project.cursor/mcp.json in the repo root
This user~/.cursor/mcp.json %USERPROFILE%\.cursor\mcp.json

Windsurf

OSConfig file
macOS / Linux~/.codeium/windsurf/mcp_config.json
Windows%USERPROFILE%\.codeium\windsurf\mcp_config.json
{
  "mcpServers": {
    "aws_kb": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-aws-kb-retrieval"
      ],
      "env": {
        "AWS_PROFILE": "mcp-readonly",
        "AWS_REGION": "us-east-1"
      }
    }
  }
}

Windsurf remote MCP: use serverUrl instead of url if the block below is HTTP.

One-liner: npx -y @modelcontextprotocol/server-aws-kb-retrieval

Secrets it wants: AWS credentials, knowledge base id

Source repo →

How to get started

  1. Create a Bedrock knowledge base and note its id.
  2. Use an IAM role that can only Retrieve on that KB.
  3. Set the standard AWS env (or a profile) plus the KB id your install expects.

Access risk

Over-broad IAM can list other KBs or invoke models you pay for. Retrieval can surface internal docs.

When to skip it

Skip it if the docs are public (Context7, Fetch) or already in a repo (Filesystem).

Instead: Context7, Fetch, Filesystem.

Vs alternatives

ServerOfficial?Needs a secret?Best for
AWS Knowledge BaseYesYesQuery an Amazon Bedrock knowledge base.
Context7NoYesUp-to-date library docs pulled into the prompt.
FetchYesYesFetch a URL and turn the page into readable text.
FilesystemYesNoRead and write files inside a folder you choose.

More in Cloud & infra

All 15 in this category →

FAQ

Is this AWS admin?

No. Retrieval. Different risk: it surfaces whatever was indexed, including things you forgot you ingested.

Official?

Anthropic reference server for AWS KB retrieve.

What gets exposed?

Chunks from the knowledge base. Treat the KB as public-to-the-model.

AWS-KB or Filesystem?

Filesystem if the files are local. KB if they were already ingested into Bedrock.