typefully
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill implements secure credential storage by saving the Typefully API key in configuration files with restricted owner-only permissions (0o600).
- [SAFE]: The skill documentation includes explicit instructions that forbid the AI agent from searching unauthorized system locations (such as the macOS Keychain, .env files, or the Trash) for credentials, instead enforcing a secure interactive setup process.
- [EXTERNAL_DOWNLOADS]: The script performs network operations targeting the official Typefully API and S3 storage for media uploads, which are necessary for its primary functionality and target well-known, trusted domains.
- [PROMPT_INJECTION]: While the skill ingests external content via the --file flag, it mitigates indirect prompt injection risks through a combination of clear automation guidelines and instructions requiring the agent to obtain user confirmation before performing public-facing or destructive actions.
- [SAFE]: The provided Node.js script is self-contained with zero external dependencies, significantly reducing the attack surface related to third-party package vulnerabilities.
Audit Metadata