The Agent Skills Directory

[SAFE]: The skill serves as a comprehensive reference and workflow guide for full-stack development. It promotes secure coding practices such as input validation, environment variable management, and proper authentication flows. No malicious behavior was identified.- [COMMAND_EXECUTION]: The skill instructions include standard CLI commands for build verification (npm run build) and local service testing (curl). These operations are typical for development environments and are restricted to local contexts.- [INDIRECT_PROMPT_INJECTION]: The skill gathers user requirements to drive architectural decisions and code scaffolding. While it lacks explicit delimiters for these inputs, the risk is minimal as the requirements are used for static code generation rather than dynamic instruction execution. * Ingestion points: User-provided technical requirements in SKILL.md (Step 0). * Boundary markers: None explicitly defined for requirement gathering. * Capability inventory: Local command execution for builds and health checks in SKILL.md (Step 4). * Sanitization: Relies on standard framework-level validation patterns (Zod, Pydantic) described in the guidelines.

fullstack-dev