The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources such as Google Search Console (GSC) and Search Engine Results Pages (SERP) to perform its diagnosis and strategy phases.
Ingestion points: External data is collected via serp_check.py and GSC insights in Phase 1, and built HTML/metadata is scanned in Phase 3.
Boundary markers: The instructions do not specify explicit delimiters to separate external data from internal agent logic.
Capability inventory: The skill performs file writes (findings, wording guides, changelogs), generates test specifications, and orchestrates deployment through repo-specific models.
Sanitization: The process relies on model reasoning and human confirmation for verdicts rather than explicit data sanitization.
[COMMAND_EXECUTION]: The skill utilizes local scripts and environment-specific tools to automate SEO checks and quality assurance.
Evidence: References to serp_check.py, independent_review.sh, and website-qa indicate dependencies on platform-provided or local scripts to perform audits and reviews.
[DYNAMIC_EXECUTION]: The skill generates regression tests based on the project's specific SEO strategy.
Evidence: Phase 3 involves writing failing specifications (trap-guards and slot-guards) derived from the WORDING-GUIDE.md to ensure site copy adheres to the new keyword strategy.

seo-reposition