swappable-crash-reporting-and-dialog

Pass

Audited by Gen Agent Trust Hub on Aug 30, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill demonstrates a design pattern for modular crash reporting. The code provided is for demonstration and architectural guidance.
  • [INDIRECT_PROMPT_INJECTION]: The skill describes an interface for handling crash data which naturally involves ingesting external strings (error messages). While this creates a potential surface for indirect injection if downstream systems are vulnerable, it is the primary intended function of the code and follows standard development practices.
  • Ingestion points: reportCrash and pushPlayerError in SKILL.md receive Throwable and PlayerError objects containing external strings.
  • Boundary markers: Not present in the architectural sample.
  • Capability inventory: Calls to a generic ReportingClient to transmit data.
  • Sanitization: Not explicitly shown in the code snippets as the data is intended for logging and reporting.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 30, 2026, 07:03 AM
Security Audit — agent-trust-hub — swappable-crash-reporting-and-dialog