The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted codebase data as its primary function. While this creates a theoretical surface for indirect prompt injection where malicious code could try to influence the auditor's output, the skill includes 'AI Self-Check' protocols and emphasizes grounding claims in local types and documentation, which serves as a mitigation.
[EXTERNAL_DOWNLOADS]: The skill recommends the installation of @ast-grep/cli via NPM and provides numerous links to research papers, blogs, and community discussions (e.g., Arxiv, Hacker News, Medium) and GitHub repositories related to code quality. These downloads and references are standard for developer tools and research-oriented skills.
[COMMAND_EXECUTION]: The auditing workflow involves running local commands such as git diff, shellcheck, ruff, and other language-specific linters. These operations are restricted to the local environment and are used as intended for static code analysis.

anti-slop