AWS OSCAL

Official Amazon Web Services Security

Lets an agent use AWS OSCAL to explore OSCAL models, retrieve schemas, and generate templates. Summary written by AI Index.

The registry’s description: AI agent tools for Open Security Controls Assessment Language (OSCAL)

Local package (pypi)

Add to your client

Add the AWS OSCAL MCP server to Claude Code

  1. Run the command

    claude mcp add aws-oscal -- uvx mcp-server-for-oscal
  2. Fill in the values ↓

  3. Check it’s connected

If the file already has servers, add the entry inside the existing object. To check it’s connected, restart or reload the client and look for aws-oscal in its list of MCP servers. A server that needs sign-in opens it on first use.

What you fill in

Required values appear as <NAME> placeholders in the snippets; add optional ones the same way. Descriptions are the server’s registry entry’s.

  • <BEDROCK_MODEL_ID> Environment variable Optional

    AWS Bedrock model ID for OSCAL documentation queries

  • <OSCAL_KB_ID> Environment variable Optional

    AWS Bedrock Knowledge Base ID for OSCAL documentation

  • <AWS_PROFILE> Environment variable Optional

    AWS CLI profile name for credential resolution

  • <AWS_REGION> Environment variable Optional

    AWS region for Bedrock API calls

  • <LOG_LEVEL> Environment variable Optional

    Logging verbosity level (DEBUG, INFO, WARNING, ERROR)

Client compatibility

By transport: whether the client’s docs list a transport this server offers. It doesn’t cover sign-in or other requirements.

Sources

This data comes from the Official MCP Registry.