skills/jeremylongshore/claude-code-plugins-plus-skills/posthog-deploy-integration/Gen Agent Trust Hub
posthog-deploy-integration
Pass
Audited by Gen Agent Trust Hub on Mar 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill facilitates application deployment via official CLI tools for trusted, well-known services including Vercel, Fly.io, and Google Cloud.
- [SAFE]: No hardcoded credentials or sensitive data are present; the documentation uses non-functional placeholders such as 'sk_live_***' for API keys to demonstrate configuration without exposing real secrets.
- [SAFE]: Deployment logic follows security best practices by utilizing platform-native secrets management commands (e.g., 'vercel secrets add', 'fly secrets set') to handle sensitive environment variables rather than hardcoding them in files.
- [SAFE]: Evaluation of indirect prompt injection surfaces (Category 8): (1) Ingestion points: Environment variables and shell parameters defined in SKILL.md. (2) Boundary markers: Absent. (3) Capability inventory: Bash tool access to vercel, fly, and gcloud CLIs. (4) Sanitization: Not explicitly defined in instructions, relying on the underlying agent platform. The implementation represents standard infrastructure-as-code patterns and does not contain malicious exploitation.
Audit Metadata