The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on executing shell commands via the feedgrab CLI tool to perform its primary functions, including fetching data, downloading media, and logging into platforms.
[INDIRECT_PROMPT_INJECTION]: The skill fetches bulk data from external, untrusted sources including X/Twitter, Reddit, Xiaohongshu, and WeChat. This content is then processed by the agent, creating a risk that malicious instructions embedded in the fetched data could influence the agent's behavior.
Ingestion points: Data returned by the feedgrab tool from various platform URLs (e.g., user timelines, search results, subreddit posts) as described throughout SKILL.md.
Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are provided in the skill instructions to guide the agent when handling the fetched content.
Capability inventory: The skill utilizes shell command execution via the feedgrab tool and performs file system operations to save outputs to directories like output/X/search/ (SKILL.md).
Sanitization: The skill does not explicitly define sanitization, validation, or filtering logic for the external content before the agent processes or summarizes it.

feedgrab-batch