opencals-build-booking-site

Pass

Audited by Gen Agent Trust Hub on Sep 22, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches official project templates from the vendor's GitHub repositories (e.g., github.com/letsopencals/template-haar) to provide a starting point for website development.
  • [COMMAND_EXECUTION]: The workflow involves running shell commands for project initialization (npx create-next-app), dependency installation (npm install), local development (npm run dev), and production deployment (npx vercel).
  • [INDIRECT_PROMPT_INJECTION]: The skill allows the agent to ingest external business descriptions and competitor site inspiration, creating a surface for indirect prompt injection.
  • Ingestion points: The skill explicitly mentions using business ideas and competitor inspiration (e.g., screenshots or URLs) to guide the build process in SKILL.md.
  • Boundary markers: There are no specific instructions or delimiters provided to isolate or ignore potential instructions hidden within the external inspiration content.
  • Capability inventory: The skill possesses significant capabilities, including file system scaffolding via npx, and network-based deployment via vercel.
  • Sanitization: No sanitization or validation mechanisms are described for the external content processed during the design phase.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 22, 2026, 11:49 PM
Security Audit — agent-trust-hub — opencals-build-booking-site