estimate-temps-savings

Warn

Audited by Gen Agent Trust Hub on Jul 21, 2026

Risk Level: MEDIUMCREDENTIALS_UNSAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill instructs the agent to search for and access sensitive environment variable files (e.g., .env, .env.example). It specifically targets high-value authentication tokens such as VERCEL_TOKEN, NETLIFY_AUTH_TOKEN, RAILWAY_TOKEN, and FLY_API_TOKEN to identify the services in use. Although the instructions include a directive to ignore secret values, accessing these files is a high-risk activity.\n- [COMMAND_EXECUTION]: The skill includes suggested shell commands for the user to install additional skills from an external repository (npx skills add gotempsh/temps --skill ...).\n- [PROMPT_INJECTION]: The skill processes untrusted project data, creating a surface for indirect prompt injection.\n
  • Ingestion points: Project manifests (package.json, requirements.txt), configuration files, and environment variable files.\n
  • Boundary markers: The skill provides instructional "Ground Rules" to delimit what information should be processed, specifically directing the agent to ignore secret values.\n
  • Capability inventory: File reading, string processing, and report generation.\n
  • Sanitization: The instructions suggest using cut to isolate environment variable keys from their values during processing.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 21, 2026, 10:03 AM
Security Audit — agent-trust-hub — estimate-temps-savings