diagnose-bugs
Pass
Audited by Gen Agent Trust Hub on Aug 21, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists solely of markdown instructions and configuration metadata. There are no scripts, binaries, or automated installation steps included in the package.
- [INDIRECT_PROMPT_INJECTION]: The instructions include a dedicated section on treating diagnostic content as untrusted. It explicitly directs the agent to ignore any command-like text, URLs, or requests found within logs, traces, or issue descriptions, effectively mitigating the risk of indirect prompt injection from external data sources.
- [DATA_EXFILTRATION]: The skill mandates that sensitive data, including credentials and customer information, must remain local by default. It provides clear rules for redacting secrets and prevents the agent from disclosing secret values during the debugging process.
- [COMMAND_EXECUTION]: The framework enforces a permission boundary, requiring explicit user authorization before the agent performs any destructive actions, modifies dependencies, installs new tools, or initiates network requests to external systems.
Audit Metadata