The Agent Skills Directory

[EXTERNAL_DOWNLOADS] (SAFE): The skill performs network requests using fetch to github.com and api.github.com. These are recognized as trusted external sources.
[PROMPT_INJECTION] (LOW): Indirect Prompt Injection Surface. The skill ingests untrusted data from external GitHub pages. 1. Ingestion points: scrapeTrending and getTrendingViaSearch functions in SKILL.md. 2. Boundary markers: None present in the code examples. 3. Capability inventory: Uses fetch to retrieve data and cheerio to parse HTML. 4. Sanitization: No sanitization or escaping is performed on the scraped repository descriptions before they are returned as data objects. This could allow a malicious repository owner to embed instructions in a description that might influence the agent if later interpolated into a prompt.

github-trending