The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external social media platforms (X, LinkedIn, Instagram, and YouTube), which represents a large surface area for untrusted input.
Ingestion points: Public posts, replies, comments, and profile data fetched via the social command line tool as described in SKILL.md.
Boundary markers: The instructions do not define specific delimiters or instructions to treat the fetched data as purely informative and ignore potential commands embedded within them.
Capability inventory: The agent has access to the bash tool, which is a high-privilege capability that could be misused if an indirect injection is successful.
Sanitization: There are no explicit sanitization or filtering steps mentioned for the fetched JSON data before it is presented to the LLM for summarization.
[EXTERNAL_DOWNLOADS]: The skill provides instructions for users to install external plugins and connect to a remote Model Context Protocol (MCP) server.
Evidence: Mentions connecting to https://mcp.hedwigai.com/ and installing the plugin /plugin marketplace add un-seen/hedwigai-skills. These resources are owned by the skill's vendor and are part of the intended setup process.

social-listening