amazon-ses
Pass
Audited by Gen Agent Trust Hub on Nov 17, 2026
Risk Level: SAFE
Full Analysis
- Robust Input Validation: The skill defines clear invariants for validating user input. It explicitly rejects values containing characters commonly used for command injection (such as backticks, semicolons, dollar signs, and quotes) and enforces strict patterns for domain names and email addresses. This effectively mitigates risks associated with dynamic shell command execution.
- Least-Privilege IAM Guidance: The skill emphasizes security by providing specific, granular lists of IAM actions required for each workflow. It explicitly forbids the use of broad wildcard permissions like
ses:*orFullAccess, encouraging the principle of least privilege. - Mandatory Data Redaction: A dedicated security consideration requires the agent to redact credentials, recipient addresses, and message content from logs or CLI output before sharing it. This practice protects user data and sensitive account information.
- Consent-Driven Operations: The skill implements mandatory consent gates for high-impact actions. For example, the agent must obtain explicit user approval before submitting production access requests, creating DNS records in Route 53, or sending real emails. This ensures the user retains control over their environment and potential costs.
- Safe Usage of Platform Tools: While the skill uses platform-specific execution tools (like
aws___run_script) for deterministic logic such as SHA-256 hashing, these are used within the scope of the platform's intended capabilities and do not involve fetching or executing untrusted external code.
Audit Metadata