The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external web content, creating an attack surface for indirect prompt injection. Ingestion points: scripts/fetch_page.py fetches remote URLs, and scripts/generate_markdown_report.py reads JSON findings. Boundary markers: SKILL.md and references/skill-contract.md both mandate treating external data as untrusted and ignoring embedded instructions. Capability inventory: Scripts perform network reads and file writes. Sanitization: scripts/fetch_page.py uses specific regex extraction for metadata, and scripts/schema_generator.py escapes values to prevent JSON injection.
[COMMAND_EXECUTION]: The skill utilizes Python scripts for core operations. scripts/fetch_page.py performs safe network requests with IP validation to prevent SSRF. scripts/generate_markdown_report.py, scripts/generate_pdf_report.py, and scripts/schema_generator.py execute file system writes to generate reports and templates as requested by the user.

seo-programmatic