The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute specific project commands such as astrofy docs check and to extract operational scripts from package.json. These actions are confined to the local development environment and serve the skill's primary purpose of documentation validation.
[SAFE]: The skill provides explicit instructions to handle environment variables securely, stating that the agent should document variable names without copying their secret values.
[SAFE]: External references point to official Astro documentation (docs.astro.build), which is a well-known and trusted resource for web developers.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a potential attack surface for indirect prompt injection because it is designed to read and process local project files, including Markdown, JSON, and source code.
Ingestion points: Reads local indices, scripts, file maps, and component consumers (defined in SKILL.md and references/technical.md).
Boundary markers: The instructions distinguish between automated checks and human annotations, requiring the agent to identify unverified claims.
Capability inventory: The agent has the ability to read project files, write documentation files, and execute the astrofy CLI tool.
Sanitization: The skill mitigates risks by requiring manual review for specific rules (manual or hybrid methods) and instructing the agent to validate all links and commands before finalizing documentation.

astrofy-documentation