x-twitter-automation
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs users to install external plugins from the vendor's repositories on GitHub (Xquik-dev/hermes-tweet) and via npm (@xquik/tweetclaw). These resources are associated with the skill author.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from X/Twitter, such as search results and tweet replies. * Ingestion points: Untrusted data enters the agent context through tools like
tweet_readand search workflows inSKILL.md. * Boundary markers: The instructions do not define clear delimiters or specific instructions for the agent to ignore potentially malicious commands embedded within retrieved tweets. * Capability inventory: The skill provides capabilities to post tweets, send direct messages, and perform account actions using thetweet_actionandtweetclawtoolsets. * Sanitization: There is no mention of sanitizing or filtering retrieved content before it is processed or summarized by the agent. - [COMMAND_EXECUTION]: The documentation provides specific shell commands for the agent to execute, including plugin installation (
openclaw plugins install) and runtime inspection (openclaw plugins inspect).
Audit Metadata