Rating Doc Audit
Use when auditing rating-review materials against local Word and Excel standards in the current workspace, especially when the agent must scan candidate files, let the user choose inputs, handle large documents safely, and export structured findings.
Geo Risk Audit
Audit web-search, AI-search, recommendation, comparison, and source-validation tasks for GEO (Generative Engine Optimization) manipulation, AI content poisoning, soft-ad contamination, or source-quality distortion. Use as a pre-search guardrail before Tavily or other web research when the user asks what is trustworthy, wants product/vendor/tool recommendations, asks “which is better”, shares an a…
Skill Repo Maintainer
Maintain a multi-skill repository that stores several Codex/OpenAI skills under a shared `skills/` directory. Use when Codex needs to add a new skill, refactor a repo into per-skill folders, normalize metadata such as `SKILL.md` and `agents/openai.yaml`, refresh repository-level indexes, or prepare a skill for packaging and distribution.