The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external social media platforms (Twitter, TikTok, Instagram, Xiaohongshu), which could contain malicious instructions designed to influence the agent. 1. Ingestion points: Content is retrieved through tools such as twitter_web_search_timeline and instagram_v3_hashtag_posts referenced in SKILL.md. 2. Boundary markers: The instructions do not provide explicit delimiters or warnings to ignore instructions embedded within the fetched posts. 3. Capability inventory: The skill uses sandbase_call_tool for data retrieval and research tasks as outlined in SKILL.md and references/sandbase-api-map.md. 4. Sanitization: No sanitization or filtering logic is described for the external content before it is processed by the agent.

hashtag-tracker