dbs
Pass
Audited by Gen Agent Trust Hub on May 19, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill logic is limited to routing user intent to other available skills. It does not perform any sensitive operations such as network requests, file system access, or command execution.
- [PROMPT_INJECTION]: The skill processes untrusted user input as its primary trigger for routing decisions, which constitutes an indirect prompt injection surface.
- Ingestion points: Captures user requests (e.g., 'help me with my business') in the main entry point to determine routing.
- Boundary markers: Absent. The skill passes user context to sub-skills without explicit delimiters or 'ignore instructions' warnings.
- Capability inventory: While this router skill has no dangerous capabilities, its routing targets (such as
/dbs-save,/dbs-restore, and/dbs-report) imply the broader toolkit has file-writing and reading permissions. - Sanitization: No sanitization of user input is performed before the routing decision is made.
Audit Metadata