omh-ralph
Pass
Audited by Gen Agent Trust Hub on May 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes arbitrary shell commands for building, testing, and linting through the
omh_gather_evidencetool. These commands are dynamically determined based on the project context to verify task completion. - [PROMPT_INJECTION]: The skill processes external plan data from files in
.omh/plans/and interpolates task descriptions into goals for subagent delegation. This creates an indirect prompt injection surface where instructions in a plan file could influence subagent behavior. - Ingestion points: Plan files located at
.omh/plans/*.mdare parsed in Step 2 to populate the task list. - Boundary markers: No explicit boundary markers or 'ignore' instructions are used when interpolating task data into the
delegate_taskgoal. - Capability inventory: The skill has access to
omh_gather_evidence(shell command execution) anddelegate_task(subagent invocation). - Sanitization: No sanitization or validation of the plan content is performed before interpolation into subagent prompts.
Audit Metadata