aws-ai

Pass

Audited by Gen Agent Trust Hub on Mar 25, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill requires cloning a repository from the author's GitHub account (github.com/KaribuLab/aws-cli-helper.git) to build a local Docker image.
  • [DATA_EXFILTRATION]: The skill accesses sensitive local files, specifically the AWS credentials directory (~/.aws) and an environment configuration file (~/.aws-ai/.env), by mounting them into a Docker container. This is a standard and necessary procedure for the skill's stated purpose of executing AWS CLI commands.
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute Docker commands to run the AWS helper. There is a naming inconsistency between the build instruction (docker build -t aws-ai) and the execution instructions (docker run ... aws_ia), which appears to be a typographical error rather than a malicious pattern.
  • [CREDENTIALS_UNSAFE]: The skill recommends storing configuration in a .env file and correctly identifies sensitive variables like AWS_AI_ASSUME_ROLE_ARN and AWS_AI_MFA_SERIAL_ARN. It explicitly instructs the agent not to expose MFA codes or session tokens in responses.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 25, 2026, 12:16 AM