debug-agent

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • Command Execution with External Data: The skill uses shell commands like az boards work-item show and gh issue view with arguments derived from external inputs such as IDs and keywords. Executing commands with variables from untrusted sources is a security consideration that requires careful handling to prevent command injection.
  • Indirect Prompt Injection Surface: The skill processes external data including pipeline logs and issue text while possessing the capability to execute commands and modify code. This configuration presents a potential surface for indirect prompt injection.
  • Ingestion points: Pipeline log files, Azure DevOps work items, and GitHub issues.
  • Boundary markers: The skill does not define specific boundary markers for the data being analyzed within the instructions.
  • Capability inventory: The skill executes CLI tools (az, gh, npm, git) and reads local files via provided tools.
  • Sanitization: The instructions do not explicitly require sanitization or validation of extracted data (e.g., IDs or search keywords) before they are interpolated into shell commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 02:58 PM
Security Audit — agent-trust-hub — debug-agent