llvm-optimization
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- Indirect Prompt Injection (LOW): The skill instructs the agent to fetch and process data from an external, untrusted GitHub repository (gmh5225/awesome-llvm-security). 1. Ingestion points: URL in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: Technical guidance on using clang and opt tools. 4. Sanitization: Absent. An attacker could potentially modify the external README to provide malicious instructions to the agent.
- External Downloads (LOW): The skill references a repository that is not part of the [TRUST-SCOPE-RULE] list. While the content is described as documentation, the source is unverifiable.
Audit Metadata