omh-curriculum-design
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
omhCLI tool (specificallyomh runtime record) to track and record the status of the curriculum design workflow. This tool is a vendor-native utility associated with the oh-my-hermes framework. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-supplied data to generate curriculum designs, creating a potential surface for indirect instructions. (1) Ingestion points: learners, learning goal, prerequisites, and constraints as defined in SKILL.md and references/procedure.md. (2) Boundary markers: No specific delimiters or safety warnings are present to isolate external inputs. (3) Capability inventory: The skill has the capability to execute the
omhCLI tool for runtime recording. (4) Sanitization: No explicit validation or sanitization of user-provided content is mentioned. Note: The risk of command injection is mitigated because the ingested data is not used as arguments for the shell commands.
Audit Metadata