omnigraph
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides comprehensive instructions for operating OmniGraph databases, including schema design, query authoring, and deployment procedures. The author (ModernRelay) and the infrastructure described (GitHub, S3, Azure) are consistent with the skill's purpose.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies the attack surface associated with processing external data such as JSONL files and schema definitions. It provides clear mitigation strategies, such as using typed parameters ($var: Type) instead of string interpolation and enforcing schema constraints (@check, @range) to ensure data integrity.
- [COMMAND_EXECUTION]: The skill uses the 'omnigraph' CLI for various graph operations. These commands are restricted to the domain-specific graph management tasks and do not expose the host system to arbitrary shell execution or privilege escalation risks.
Audit Metadata