break-ui
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze local source code and schema files. This represents an attack surface for indirect prompt injection if those files contain malicious instructions.
- Ingestion points: Reads UI component files and schema files (Zod, Yup, etc.) as described in SKILL.md Phase 1.
- Boundary markers: Includes an explicit safety rule (Hard Rule 6) instructing the agent to treat repository content as data only and to flag any "ignore previous instructions" patterns.
- Capability inventory: The skill can modify component files to add data fixtures and toggle controls (Phase 3 and 6).
- Sanitization: Instructions specify that any "steering" attempts found in files should be flagged rather than followed.
- [PROMPT_INJECTION]: While the skill contains the phrase "ignore previous instructions", it is used as a negative example within a safety rule (Hard Rule 6). It instructs the agent on what to detect and ignore in external data, rather than being an attempt to subvert the agent's own behavior.
Audit Metadata