z-qa-debugger
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface as it instructs the agent to analyze external test runner outputs and log files which could theoretically be manipulated to include instructions.
- Ingestion points: The agent reads command outputs, stack traces, and test evidence files such as logs and
QA_REPORT.md(SKILL.md). - Boundary markers: The instructions do not define specific delimiters or warnings to ignore instructions embedded within test logs.
- Capability inventory: The skill authorizes the agent to execute shell commands using various test runners (Go, Cargo, pytest, Vitest, Playwright) and container management tools like Docker and Kubernetes (SKILL.md).
- Sanitization: No explicit sanitization or filtering of log content is mandated prior to processing.
Audit Metadata