The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses local Node.js scripts located in the scripts directory to perform authentication checks, documentation resolution, and live API requests.
[CREDENTIALS_UNSAFE]: The script 'scripts/configure-keyapi-auth.mjs' modifies user shell profiles (such as .bashrc, .zshrc, .profile, and PowerShell profiles) to persist the KEYAPI_TOKEN environment variable. While this is used for setup, modifying shell startup files is a sensitive operation that can expose credentials or be used for persistence.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from YouTube, such as video metadata and user comments, without explicit sanitization. Ingestion points: YouTube API responses (videos, comments, search results) processed in SKILL.md and through scripts. Boundary markers: The skill lacks explicit delimiters or instructions to ignore embedded commands in the external data. Capability inventory: Includes file system writes (via configure-keyapi-auth.mjs) and network operations (via keyapi-api.mjs). Sanitization: No specific validation or filtering of the YouTube content is performed before presentation to the agent.
[EXTERNAL_DOWNLOADS]: The skill dynamically fetches API contract information and endpoint documentation from 'docs.keyapi.ai' at runtime to construct REST requests.

keyapi-youtube