The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes various local Python scripts (e.g., google_auth.py, backlinks_auth.py, drift_history.py, google_report.py) to handle authentication, track history, and generate PDF deliverables.
[COMMAND_EXECUTION]: It utilizes command-line tools such as curl and npx lighthouse to perform real-time performance auditing and fetch raw data from external endpoints.
[EXTERNAL_DOWNLOADS]: The skill interacts with numerous well-known technology and data providers, including Google APIs (Search Console, PageSpeed, CrUX), Moz, Bing, OpenStreetMap (Overpass and Nominatim), and specialized services like DataForSEO and Geoapify.
[PROMPT_INJECTION]: As a tool designed to crawl and analyze external websites, the skill possesses an indirect prompt injection attack surface. Content from untrusted, user-provided URLs is ingested into the agent's context for evaluation.
Ingestion points: Crawled page content, metadata, and schema markup from audited URLs.
Boundary markers: No explicit delimiters or 'ignore instructions' warnings are mentioned for the ingestion of crawled content.
Capability inventory: The skill has access to shell execution via Python scripts and network operations via curl.
Sanitization: No specific sanitization or filtering of external content is documented.

seo