The Agent Skills Directory

[SAFE]: The skill's primary function is to enforce rigorous software engineering standards. It provides a logical framework (Gate functions) to guide the agent toward secure and maintainable code practices, such as validating data at boundaries and avoiding silent error failures.
[INDIRECT_PROMPT_INJECTION]: The skill is intended to be used during code reviews and debugging of external, untrusted code, which is a standard surface for indirect prompt injection. However, the skill's own instructions explicitly advocate for validation at boundaries, which serves as a mitigation against such attacks. * Ingestion points: Software source code, pull request descriptions, and bug reports supplied by the user during engineering tasks. * Boundary markers: Not explicitly defined within the skill instructions, though the skill encourages the creation of validation boundaries in the code it processes. * Capability inventory: Code analysis, modification, and architectural decision-making across the project codebase. * Sanitization: The skill does not provide specific sanitization routines but mandates the use of schema validation (e.g., Zod) for any incoming external data in the code it generates.

no-workarounds