tooluniverse-publish-tool
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest metadata from remote tools using
list_remote_toolsand then pass that information (name, description, inputs) into theprepare_tool_publicationtool. This creates a potential attack surface where a maliciously crafted tool name or description on the user's computer could influence the agent's behavior or the parameters of the publication call. - Ingestion points: The agent retrieves tool metadata (name, description, input schema) from the output of the
list_remote_toolscommand. - Boundary markers: The instructions do not specify the use of delimiters, escaping, or strict boundary markers when the agent incorporates the retrieved metadata into the arguments for the
prepare_tool_publicationtool call. - Capability inventory: The skill utilizes
list_remote_tools(read),run_remote_tool(execute), andprepare_tool_publication(write/create draft). - Sanitization: The skill lacks explicit instructions for the agent to sanitize, validate, or filter the retrieved external tool metadata before processing it.
Audit Metadata