The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from Instagram, which can serve as a vector for indirect prompt injection attacks.
Ingestion points: The skill reads external content through instagram-cli and hatch-zeitgeist, including post captions, reels, comments, and user profile biographies (documented in SKILL.md and references/content-questions.md).
Boundary markers: The instructions do not define specific delimiters or security headers to separate untrusted external content from agent instructions during processing.
Capability inventory: The skill possesses significant capabilities, including the ability to write to the user's account (post-feed, post-story, update-bio), access the local file system (~/workspace/), and spawn browser tasks for web inspection.
Sanitization: There is no evidence of content sanitization, escaping, or filtering before external data is interpolated into the agent's context.
[COMMAND_EXECUTION]: The skill relies on the execution of several CLI tools to perform its tasks.
Evidence: It uses exec to run instagram-cli for account management and data retrieval. It also utilizes hatch-zeitgeist for enriched data lookups and ffmpeg for media processing, as described in references/content-questions.md.

instagram