omh-meta-router

Pass

Audited by Gen Agent Trust Hub on Sep 25, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the omh command-line utility to query a workflow catalog (e.g., omh recommend "<remainder>" --json --limit 3). This tool is central to the skill's routing function and is a vendor-native utility.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input following the /omh trigger. This data is passed to the omh CLI. The skill attempts to mitigate risks by using structured JSON output, limiting the number of results, and maintaining an evidence boundary. 1. Ingestion points: User command remainder in SKILL.md. 2. Boundary markers: Instructions to use observed-vs-prepared evidence boundaries. 3. Capability inventory: Command execution via the omh CLI. 4. Sanitization: Use of --json and --limit flags in the recommendation query.
  • [DYNAMIC_EXECUTION]: The skill performs dynamic workflow selection and chaining based on output from catalog queries. This behavior is the primary intended function of the meta-router and relies on structured metadata from the recommendation system.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 25, 2026, 08:34 PM
Security Audit — agent-trust-hub — omh-meta-router