The Agent Skills Directory

[SAFE]: The skill primarily functions by fetching and analyzing web content via WebFetch to audit SEO freshness signals. This behavior is directly aligned with its stated purpose.
[SAFE]: The skill processes untrusted external HTML data, which presents an indirect prompt injection surface. Ingestion points: rendered_dom and raw_html from external URLs fetched via WebFetch. Boundary markers: Absent. Capability inventory: Access to Bash, WebFetch, Read, Grep, and Glob. Sanitization: Not explicitly mentioned. This risk is inherent to web auditing tools and is handled here safely for the intended purpose.
[SAFE]: No hardcoded credentials, sensitive file access patterns, or data exfiltration behaviors were identified.
[SAFE]: References to internal scripts, such as scripts/check-freshness.mjs, are used for diagnostic and verification purposes and do not involve downloading or executing code from untrusted remote sources.
[SAFE]: The skill includes explicit instructions for the agent to maintain "Honesty" and avoid fabricating data, which is a positive safety constraint.

seo-freshness