test-evidence-review
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill performs read-only analysis of local project files (markdown stories, test source code, and evidence documents). It does not engage in remote code execution, external downloads, or data exfiltration.
- [SAFE]: File system interactions are limited to standard
Read,Glob,Grep, andWriteoperations for local documentation and source files within the project structure (e.g.,production/epics/,tests/unit/). - [SAFE]: While the skill can write a report file (
production/qa/evidence-review-[date].md), it is configured to ask for user confirmation before doing so, following standard development practices. - [SAFE]: No prompt injection or obfuscation patterns were detected. The skill uses clearly defined heuristics for quality assessment (e.g., counting assertions, checking for specific keywords like 'null' or 'edge') that do not involve executing untrusted data.
Audit Metadata