pattern-test
Warn
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- COMMAND_EXECUTION (LOW): The skill utilizes 'deno task ct test' to run tests, which is an execution capability. While legitimate for the skill's purpose, it provides a vector for code running in the local environment.\n- PROMPT_INJECTION (MEDIUM): The skill possesses an Indirect Prompt Injection surface (Category 8). Ingestion points: Reads and tests code from 'packages/patterns/'. Boundary markers: Lacks markers to separate instructions from code logic. Capability inventory: Command execution via Deno. Sanitization: No sanitization of code within the patterns is performed. A malicious pattern could contain instructions or code that compromises the agent's context or the execution environment.
Audit Metadata