omh-cto-loop
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external "operating signals" and "roadmap scope" to generate technical tradeoffs and decisions, creating a surface for indirect prompt injection.
- Ingestion points: The skill requires
operating signalsandroadmap or release scopeas inputs from the environment or user (SKILL.md). - Boundary markers: There are no explicit instructions to the agent to treat input signals as data only or to ignore instructions embedded within them.
- Capability inventory: The skill utilizes the
omhCLI tool and theskill_viewfunction to manage workflow and reference external guides (SKILL.md). - Sanitization: The instructions do not specify any validation or filtering for the input signals.
- [COMMAND_EXECUTION]: The skill uses a vendor-specific command-line tool
omhto record workflow status. - Evidence:
omh runtime record --skill cto-loop --harness app-delivery-loop --status started(SKILL.md) is used to track the progress of the leadership loop.
Audit Metadata