omh-application-threat-model
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or dangerous command executions were detected. The skill is purely instructional and provides a methodology for security review without using risky tools or external dependencies.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input regarding system architecture and data flows. 1. Ingestion points: System components, data flows, and trust boundary descriptions provided by the user in SKILL.md. 2. Boundary markers: No specific boundary markers or delimiters for user input are defined. 3. Capability inventory: The skill performs analysis and documentation only; no tools for filesystem writes, network operations, or shell execution are invoked. 4. Sanitization: The instructions explicitly forbid the generation of functional exploit code, payloads, or the printing of secrets and live records, focusing instead on preconditions and detection signals.
Audit Metadata