The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The documentation includes a link to installation instructions for the vendor's CLI tool hosted on their official GitHub repository.
[INDIRECT_PROMPT_INJECTION]: The skill defines an attack surface for indirect prompt injection through its use of the infsh/agent-browser tool.
Ingestion points: Untrusted data enters the agent context via the infsh/agent-browser command used to visit external URLs.
Boundary markers: The instructions do not define specific boundary markers or 'ignore' instructions for data processed from the browser.
Capability inventory: The skill possesses capabilities to execute shell commands via belt, post content to social media (x/post-create), and perform web searches (tavily/search-assistant).
Sanitization: There is no evidence of automated sanitization or filtering of content retrieved from external web pages before it is processed by the agent.
[COMMAND_EXECUTION]: The skill lists several commands using the belt CLI tool for social media management and media generation. These operations are within the scope of the skill's intended purpose and are authorized via the allowed-tools configuration in the manifest.

twitter-thread-creation