reviewing-design-work
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [SAFE]: The skill serves as a high-level orientation and workflow dispatcher for design reviews. It does not contain scripts, shell commands, network operations, or sensitive file access.
- [INDIRECT_PROMPT_INJECTION]: This skill defines a workflow for processing external, untrusted data including pull request diffs, UI screenshots, and Figma specifications. Ingestion points include these design surfaces and code diffs. There are no boundary markers or sanitization logic present to separate data from instructions. However, the skill's capability inventory shows a lack of dangerous operations (such as file-system writes or network exfiltration), making the injection surface non-exploitable by this specific skill.
- [EXTERNAL_DOWNLOADS]: The skill metadata references a source repository at github.com/thrillmade/agent-skills. This resource is associated with the verified author and does not involve unauthorized remote code execution or untrusted package installation.
Audit Metadata