axiom-ios-integration
Pass
Audited by Gen Agent Trust Hub on Feb 18, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [Indirect Prompt Injection] (SAFE): The skill defines a routing table for user queries but does not execute logic that could be subverted via the router itself.
- Ingestion points: User-provided natural language queries regarding iOS features.
- Boundary markers: Not applicable, as this is a high-level router that delegates to other skills.
- Capability inventory: Limited to internal skill redirection (
/skill) and agent handoff. - Sanitization: Not present, as it acts as a lookup table.
- [Prompt Injection] (SAFE): No override markers or jailbreak-style instructions detected. The instructions are purely functional for task routing.
- [Data Exposure & Exfiltration] (SAFE): No access to sensitive system paths or external network requests identified.
- [Remote Code Execution] (SAFE): No remote script downloads or dynamic execution patterns found.
Audit Metadata