ue-actor-component-architecture
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill suggests reading external project context from '.agents/ue-project-context.md' to adapt its advice to the user's specific project setup. While this provides project-specific utility, it introduces a potential surface for indirect prompt injection if the context file contains malicious instructions.
- Ingestion points: '.agents/ue-project-context.md' referenced in SKILL.md.
- Boundary markers: Absent.
- Capability inventory: The skill is primarily instructional for C++ development and does not include active execution capabilities like network or shell access.
- Sanitization: Absent.
Audit Metadata