The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates the retrieval of external data from Twitter/X, which creates a surface for instructions embedded in social media content to influence the agent's actions.
Ingestion points: Operations like x/post-get and x/user-get (found in SKILL.md) fetch content from a public, attacker-controllable platform.
Boundary markers: The instructions do not define clear delimiters or "ignore instructions" wrappers for the data fetched from Twitter.
Capability inventory: The agent is granted the ability to perform high-impact actions on behalf of the user, including x/post-tweet, x/post-delete, x/dm-send, and x/user-follow using the belt CLI tool.
Sanitization: There is no evidence of content filtering or sanitization being applied to the fetched tweet data before it is presented to the agent.
[EXTERNAL_DOWNLOADS]: The skill relies on external components for its core functionality, including the belt-sh/cli skill and installation scripts hosted at github.com/inference-sh/skills.

twitter-automation