The Agent Skills Directory

[SAFE]: The skill defines a standard for machine-readable documentation indices and provides implementation examples for various web frameworks.
[INDIRECT_PROMPT_INJECTION]: The skill describes processes for aggregating documentation files into indices for LLM consumption. Evidence chain: (1) Ingestion points: documentation files located in docs/, content/, and src/pages/ as described in the CI automation section of SKILL.md. (2) Boundary markers: Recommends using # File: filename and --- separators in generated companion files. (3) Capability inventory: Utilizes write_file, create_file, and fetch tools to generate and verify manifest files. (4) Sanitization: Specifically includes OWASP ASI guardrails, such as mandatory schema validation (ASI05) and link integrity checks (ASI04).
[COMMAND_EXECUTION]: Provides standard example scripts for file concatenation and GitHub Actions workflows which are routine for static site generation and CI/CD automation.

configure-llms-txt