The Agent Skills Directory

[COMMAND_EXECUTION]: Executes shell commands to install system dependencies and fetch external data.
In references/newsjacking.md, it runs brew install jq to ensure JSON processing capabilities are available.
Uses curl to query APIs and RSS feeds from well-known services including Google, Algolia, and Reddit.
[EXTERNAL_DOWNLOADS]: Retrieves data from well-known external services to identify trending news.
Downloads RSS feeds from news.google.com.
Fetches JSON data from the Hacker News Algolia API and Reddit's JSON endpoints.
[PROMPT_INJECTION]: Potential for indirect prompt injection due to processing untrusted external data.
Ingestion points: Fetches content from public news feeds and social media profiles via curl and dev-browser in references/newsjacking.md and references/journalist-pitching.md.
Boundary markers: No delimiters or "ignore instructions" warnings are provided to the agent when processing this untrusted external data.
Capability inventory: The agent can execute shell commands (curl, jq, brew), use a browser to read arbitrary URLs, and write to local files (.agents/media-list.md).
Sanitization: The skill does not implement validation, filtering, or escaping of the content retrieved from external sources before it is interpreted by the agent.

public-relations