z-debugging

Pass

Audited by Gen Agent Trust Hub on Apr 22, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines a logical debugging workflow including environment isolation using git worktrees and log analysis using standard tools such as Docker and .NET Aspire. These operations are consistent with the skill's stated purpose.
  • [PROMPT_INJECTION]: The skill instructions involve collecting user-provided context to drive the debugging process (e.g., creating a git worktree slug based on the issue summary). This represents a surface for indirect prompt injection common to debugging agents, but no malicious intent was found.
  • Ingestion points: User communication via ask/question tools and issue descriptions used for creating worktree slugs.
  • Boundary markers: Absent in this specific instruction file.
  • Capability inventory: Git operations, Docker/Aspire log viewing, and tool execution via referenced skills (z-aspire-orchestrator).
  • Sanitization: Not explicitly mentioned, relying on the platform's tool execution constraints.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 22, 2026, 11:53 AM
Security Audit — agent-trust-hub — z-debugging