ontology-grounded-spec
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious code or patterns were detected. The skill defines a standard workflow for requirement engineering and documentation.- [COMMAND_EXECUTION]: The skill utilizes project-specific CLI tools (lp task, lp registry) to manage task states and project registries during the specification process.- [INDIRECT_PROMPT_INJECTION]: The skill ingests user input to generate technical specifications. It mitigates potential risks by instructing the agent to treat ambiguous user statements as hypotheses requiring verification rather than direct instructions, ensuring human-in-the-loop validation of intent.
Audit Metadata