The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The script scripts/seo_audit.py uses the urllib library to fetch content from external URLs provided by the user. This is a functional requirement to analyze page titles, descriptions, and accessibility for AI search engines.\n- [COMMAND_EXECUTION]: The scripts/verify.sh bash script executes grep or ripgrep to perform static analysis of project files. It identifies prohibited marketing terminology and verifies basic structural requirements like the presence of a single H1 tag.\n- [PROMPT_INJECTION]: The skill facilitates the ingestion of external content, which represents a surface for indirect prompt injection.\n
Ingestion points: Raw writing samples provided by the user are stored in 02-DOCS/raw/brand/, and metadata is retrieved from external websites via scripts/seo_audit.py.\n
Boundary markers: Absent. The skill instructions do not specify the use of delimiters or warnings when the agent processes these raw text samples.\n
Capability inventory: The skill performs network requests through scripts/seo_audit.py, executes local search commands via scripts/verify.sh, and manages project documentation by writing to the 02-DOCS/ directory and the CLAUDE.md file.\n
Sanitization: Absent. Content extracted from external sources is processed as plain text without specialized filtering for instructions designed to influence the agent's behavior.

marketing