The Agent Skills Directory

[PROMPT_INJECTION]: The skill analyzes untrusted project data (codebase files) to generate audit reports and site artifacts, which creates an indirect prompt injection surface.
Ingestion points: The skill reads various project source files including HTML, TSX, Markdown, and JSON files using Read, Glob, and Grep tools during its detection and audit phases.
Boundary markers: There are no explicit instructions or delimiters provided to the agent to treat the content of audited files as data or to ignore instructions that might be embedded within them.
Capability inventory: The skill has the authority to use Write and Edit tools to create or modify critical files like robots.txt, llms.txt, and schema JSON-LD templates based on analyzed data.
Sanitization: No specific sanitization or validation logic is defined to ensure content extracted from the project files is safe before being interpolated into audit results or generated site artifacts.
[EXTERNAL_DOWNLOADS]: The skill references several external resources and validation tools as part of its documentation and reporting.
References llm-txt-validator.vercel.app (Vercel) and Google's Rich Results Test for manual validation of generated artifacts.
Points to reputable academic research repositories (arXiv.org) and practitioner websites for deeper context on its scoring rubric.
These references are informational and do not involve automated script execution or package installation by the skill itself.

geo-seo