alien-agent-id

Pass

Audited by Gen Agent Trust Hub on May 13, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes node and git commands locally to manage cryptographic keys and sign git commits. These operations are scoped to the agent's state directory (~/.agent-id) and standard git workflows.
  • [EXTERNAL_DOWNLOADS]: The skill fetches OIDC configuration and service manifests from remote URLs. These interactions are protected by strict schema validation, size limits (8KB), and authority-matching rules to ensure the agent only interacts with the intended service infrastructure.
  • [CREDENTIALS_UNSAFE]: While the skill handles sensitive credentials, it implements an encrypted local vault using AES-256-GCM with keys derived via HKDF from the agent's private key. The documentation actively discourages insecure practices, such as passing secrets via command-line arguments, and provides secure alternatives like environment variables and file-based inputs.
Audit Metadata
Risk Level
SAFE
Analyzed
May 13, 2026, 04:52 PM
Security Audit — agent-trust-hub — alien-agent-id