repo-delivery
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides process-oriented guidelines for repository maintenance and onboarding. It contains no executable scripts, binaries, or automated command sequences.- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for interacting with external repository content which represents a potential attack surface.
- Ingestion points: Repository checkout, package metadata, release history, and documentation (SKILL.md).
- Boundary markers: None specified.
- Capability inventory: Validation commands and deterministic checks mentioned as part of the verification process.
- Sanitization: Not specified.- [SAFE]: The instructions explicitly advise against dangerous behaviors, such as silently changing machine policies or claiming local builds as official releases, and emphasize the isolation of test credentials.
Audit Metadata