The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes shell command snippets using curl, jq, and xmllint. These are used as utility examples for the agent to fetch and process data from public RSS feeds and APIs for research purposes.
Evidence: Found in references/newsjacking.md (for Google News, Hacker News, and Reddit) and references/journalist-pitching.md.
[EXTERNAL_DOWNLOADS]: The skill facilitates the retrieval of content from external domains to assist with PR research, such as monitoring industry trends and journalist activity.
Evidence: Reference files provide URLs and command structures for fetching data from well-known technology and news services.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external data, such as news articles and podcast descriptions, which could theoretically contain adversarial content.
Ingestion points: RSS feeds and website scraping in references/newsjacking.md and references/podcast-guest-prep.md.
Boundary markers: The instructions do not define specific delimiters for external content.
Capability inventory: Data fetching and summarization for generating pitches and briefs.
Sanitization: No explicit sanitization or validation of the fetched content is provided.

public-relations