The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes several Python utility scripts (fetch_page.py, generate_markdown_report.py, schema_generator.py, seo_health_score.py) intended to be used as local tools for data processing and report generation.
[EXTERNAL_DOWNLOADS]: The generate_pdf_report.py utility suggests the manual installation of the reportlab library via pip if the user requires optional PDF rendering functionality.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function involves fetching and processing external web content, which creates a surface for indirect prompt injection.
Ingestion points: The fetch_page.py script retrieves content from external URLs, and the agent reads user-provided project files.
Boundary markers: The SKILL.md instructions require the agent to label evidence sources (e.g., measured, user-provided) but do not define specific technical delimiters for the ingested data.
Capability inventory: The skill has network access via fetch_page.py and file-writing capabilities via generate_markdown_report.py, generate_pdf_report.py, and schema_generator.py.
Sanitization: The references/skill-contract.md file explicitly mandates that scraped material must be treated as untrusted input and that any instructions embedded within that data must be ignored.
[DATA_EXFILTRATION]: The fetch_page.py script performs outbound network requests to retrieve SEO data. It implements Server-Side Request Forgery (SSRF) protections by validating that hostnames do not resolve to private, loopback, or reserved IP addresses.

keyword-research