The Agent Skills Directory

[Remote Code Execution] (CRITICAL): The skill directs users to execute curl -fsSL https://cli.inference.sh | sh. This is a critical security risk because it executes a remote script directly in the shell without any verification or code signing checks. The domain inference.sh is not a trusted source.
[External Downloads] (MEDIUM): The skill uses npx skills add to fetch and install external dependencies from inference-sh/skills. Since this repository and organization are not in the trusted list, this introduces unverified code into the user's environment.
[Indirect Prompt Injection] (LOW): The skill has a surface for indirect prompt injection via external search results.
Ingestion points: Data retrieved from tavily/search-assistant is used as research for generating posts.
Boundary markers: No delimiters are used to separate the external search content from the agent's instructions.
Capability inventory: The skill possesses the Bash tool with infsh capabilities, enabling it to perform network operations and post to external accounts.
Sanitization: The instructions do not specify any sanitization or filtering of the search results before they are processed by the LLM.

linkedin-content