The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches the @xquik/tweetclaw plugin and the mppx and viem libraries from the npm registry. These are standard dependencies for the skill's stated purpose of X automation and blockchain-based metered billing.
[COMMAND_EXECUTION]: Uses the openclaw command-line utility for plugin installation and local configuration management.
[SAFE]: Follows security best practices for credential handling by instructing users to use environment variables ($XQUIK_API_KEY, $MPP_SIGNING_KEY) and warning against hardcoding secrets.
[SAFE]: Implements a robust safety policy that requires explicit user approval for all 'visible actions' such as posting, replying, following, or sending DMs.
[PROMPT_INJECTION]: The skill handles untrusted data from the X platform (tweets, replies, profiles).
Ingestion points: The tweetclaw tool retrieves external tweet content and user data.
Boundary markers: Not explicitly defined for the content of ingested tweets.
Capability inventory: The skill can perform network operations via API calls and execute social actions (Post, DM, Follow).
Sanitization: Not specified for ingested content.
Mitigation: The risk of indirect prompt injection is mitigated by the mandatory approval-gated workflow for all state-changing actions.

tweetclaw