IAM Policy Autopilot Logo

IAM Policy Autopilot

Official

Analyzes application code locally to automatically generate baseline AWS IAM identity-based policies...

7 views
0 installs
Updated Jan 20, 2026
Not audited
Analyzes application code locally to automatically generate baseline AWS IAM identity-based policies by detecting AWS SDK calls in Python, Go, and TypeScript applications. Helps AI coding assistants quickly create IAM permissions that can be refined as applications evolve.
  1. Create or confirm AWS credentials you will use

    • If you already have an IAM user with programmatic access, skip to step 2.
    • To create new programmatic credentials in the AWS Console:
      1. Sign in to the AWS Management Console and open the IAM console.
      2. Go to "Users" → "Add user".
      3. Enter a username, check "Programmatic access", attach appropriate policies (or add to a group with the required permissions), and finish creation.
      4. Copy the Access key ID and Secret access key (you will need these once).
  2. (Recommended) Create an AWS CLI profile (so you can use AWS_PROFILE)

    • Install and configure the AWS CLI if needed: run aws configure --profile my-profile-name and enter:
      • AWS Access Key ID
      • AWS Secret Access Key
      • Default region name (e.g., us-east-1)
      • Default output format (optional)
    • Confirm the profile exists in ~/.aws/credentials (or C:\Users\<you>\.aws\credentials on Windows).
  3. Decide which environment variables you will provide to the MCP server

    • Preferred method shown in the README: set
      • AWS_PROFILE — the name of the AWS CLI profile you created (e.g., my-profile-name)
      • AWS_REGION — the AWS region to use (e.g., us-east-1)
    • Alternative (if you cannot use an AWS profile): set these directly instead of AWS_PROFILE:
      • AWS_ACCESS_KEY_ID
      • AWS_SECRET_ACCESS_KEY
      • (If using temporary credentials) AWS_SESSION_TOKEN
  4. Determine the exact values to enter

    • AWS_PROFILE: the profile name you used with aws configure (or an existing profile name from ~/.aws/credentials).
    • AWS_REGION: the AWS region you want the tool to assume (e.g., us-east-1, us-west-2). Use the region that matches your resources or us-east-1 if unsure.
    • AWS_ACCESS_KEY_ID / AWS_SECRET_ACCESS_KEY / AWS_SESSION_TOKEN: copy these exactly from IAM console when you created the access keys (secret is shown only once).
  5. Open the FastMCP connection interface and click the "Install Now" button

    • In the Install Now dialog (or the form that appears), add the environment keys and paste their values exactly as determined above.
  6. What to put in each FastMCP ENV field (examples)

    • AWS_PROFILE: my-profile-name
    • AWS_REGION: us-east-1
    • OR (if not using profile)
      • AWS_ACCESS_KEY_ID: AKIA...
      • AWS_SECRET_ACCESS_KEY: wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY
      • AWS_SESSION_TOKEN: (only for temporary creds)
  7. Save/apply the FastMCP settings and start the MCP server

    • After you click the Install/Save button, confirm the MCP server logs show it picked up the environment values (or that the connection succeeded).
  8. Verify the integration

    • Run a simple IAM Policy Autopilot command via the MCP or local CLI (for example iam-policy-autopilot mcp-server --transport http) or ask your AI coding assistant to generate a policy to ensure the server can access AWS as expected.
    • If you used AWS_PROFILE, ensure the machine running the MCP can read ~/.aws/credentials for that profile; if not, use direct access key environment variables instead.
  9. Security reminder

    • Do not share AWS secret keys. Use IAM least-privilege policies and prefer named CLI profiles or temporary credentials where possible.

Quick Start

View on GitHub

More for Security

View All →

More for Developer Tools

View All →

Report Issue

Thank you! Your issue report has been submitted successfully.