The Agent Skills Directory

[SAFE]: The skill and its reference files contain no evidence of malicious intent, prompt injection, or unauthorized data access. The content is consistently focused on improving application security posture through structured auditing and best-practice implementation.
[INDIRECT_PROMPT_INJECTION]: This skill intentionally ingests untrusted user content (source code and configurations) for the purpose of security auditing. While this provides a surface for indirect prompt injection, the skill's internal logic manages the associated risks. Ingestion points: application code and config files provided by the user during the audit process. Boundary markers: the skill requires findings to be documented in a specific format with verifiable evidence (test names, config lines). Capability inventory: the skill is restricted to analyzing code and providing fix recommendations. Sanitization: instructions explicitly forbid the use of destructive payloads or real data extraction against production environments.

security-baseline