The Agent Skills Directory

[SAFE]: Detailed analysis of the skill files revealed no evidence of prompt injection, data exfiltration, or malicious obfuscation. The skill is designed for static analysis of SEO metadata and content strategy.
[INDIRECT_PROMPT_INJECTION]: The skill evaluates external content (such as web pages and project files) using validation rules.
Ingestion points: Processes user-provided source files (HTML, TSX, JSX, MD) to validate SEO markers like meta tags and keyword density.
Boundary markers: The skill instructs the agent to ground all responses in the provided reference system files.
Capability inventory: No dangerous capabilities such as unconstrained network access or file system write permissions are present.
Sanitization: Uses predefined regex patterns for validation; does not interpolate untrusted data into sensitive execution sinks.
[COMMAND_EXECUTION]: The documentation includes bash and SQL snippets (e.g., using grep, sed, and npx @lhci/cli). These are presented as instructional examples for the user to facilitate technical SEO audits and are not intended for automated execution by the agent platform.

seo