The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The trend-discovery module ingests untrusted external data by scanning the Google News RSS feed (news.google.com). This presents a vulnerability surface where malicious actors could attempt to influence the agent's trend discovery and briefing process via crafted news headlines.
Ingestion points: modules/trend-discovery/MODULE.md (Google News RSS search query).
Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands within the fetched RSS items.
Capability inventory: The agent can load transcripts, search project content, and publish social posts via the Fountain API. It also has instructions allowing for script generation and local file storage.
Sanitization: There is no evidence of sanitization or filtering for the external headlines and descriptions before they are used to generate search terms or briefs.
[DYNAMIC_EXECUTION]: The HOUSEKEEPING.md file provides instructions allowing the agent to write and execute "throwaway scripts" to handle repetitive API requests or large data volumes. While intended for efficiency, this involves runtime generation and execution of executable content.
[EXTERNAL_DOWNLOADS]: The skill uses curl to fetch information from Google News (news.google.com). This is a well-known service and the usage is consistent with the skill's primary purpose of discovering current trends.

fountain-daily-growth