The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references installation instructions and sub-skills hosted on the vendor's official GitHub repository and infrastructure.
Evidence: https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md and npx skills add commands for vendor-owned skills.
[COMMAND_EXECUTION]: The skill uses the belt CLI tool to perform all actions, including posting tweets, sending direct messages, and managing user profiles. The use of the tool is restricted via the frontmatter to the specific CLI binary.
Evidence: Use of belt app run for various Twitter operations.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from an external source (Twitter/X) through tools that retrieve tweets and user profiles. This data is then available in the agent's context and could influence subsequent actions.
Ingestion points: x/post-get and x/user-get retrieve external content.
Boundary markers: None explicitly defined in the skill instructions.
Capability inventory: The skill can post new content (x/post-tweet), send messages (x/dm-send), and follow users (x/user-follow) via the belt CLI.
Sanitization: No specific sanitization or filtering of the retrieved social media content is mentioned.

twitter-automation