xstate
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides documentation and reference materials for using XState v5. Analysis across all threat categories, including prompt injection, data exfiltration, and remote code execution, confirms the absence of malicious patterns.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a clear attack surface by instructing the agent to process data from external sources, but it mandates defensive measures.
- Ingestion points: External events from sources like WebSockets, storage, and URLs are identified as entry points for untrusted data in
typescript-and-migration.md. - Boundary markers: The instructions require schema-first parsing and validation before events are processed by the state machine.
- Capability inventory: No risky system-level capabilities or unsafe code execution patterns are introduced or recommended for handling untrusted data.
- Sanitization: The skill mandates strict validation at all trust boundaries as part of the core modeling workflow.
Audit Metadata