The Agent Skills Directory

[SAFE]: The skill provides a template for a developer-oriented autonomous pipeline. It references a local workflow instruction file located at ~/.claude/gsd-core/workflows/audit-fix.md and leverages standard development tools (Bash, Edit, Write) for its stated purpose. No obfuscation, prompt injection, or suspicious network activities were found.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from code audit reports to perform modifications, which constitutes an ingestion surface for potentially untrusted data. This surface is integral to the skill's primary function. * Ingestion points: Output from audit commands specified via the --source parameter. * Boundary markers: None explicitly defined in this wrapper. * Capability inventory: Bash, Write, Edit, and Agent tools. * Sanitization: None specified in the current instruction set.

gsd-audit-fix