Aphorisms
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection via the
FindAphorismworkflow. \n - Ingestion points: The skill uses the
WebFetchtool to retrieve content from arbitrary user-provided URLs for thematic analysis inWorkflows/FindAphorism.md. \n - Boundary markers: No explicit delimiters or safety instructions are defined to ensure the agent ignores potential commands embedded within the fetched external content. \n
- Capability inventory: The skill has the ability to write to the local file system using the
Edittool and initiate network requests viacurl. \n - Sanitization: There is no evidence of content sanitization or validation for the data retrieved from external sources before it is processed by the reasoning engine. \n- [COMMAND_EXECUTION]: The skill mandates the execution of a
curlcommand through a shell subprocess upon every invocation to send a status notification to a local service (localhost:8888). While the destination is restricted to the local host, the practice of executing shell commands with agent-interpolated strings increases the attack surface.
Audit Metadata