ext-triage-java

Pass

Audited by Gen Agent Trust Hub on Jul 11, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [NO_CODE]: The skill is composed strictly of markdown documentation files used for knowledge-based decision making. It does not contain any scripts, binaries, or executable logic.
  • [SAFE]: No network operations, hardcoded credentials, or sensitive file access patterns were detected. All external references are to internal skills or established development documentation.
  • [SAFE]: The skill defines a surface for indirect prompt injection as it triages findings from external tools and PR comments. However, it incorporates multi-layered safety checks, including a plan-intent validation step and a requirement for user escalation for ambiguous or high-risk suggestions.
  • Ingestion points: PR review comments, Sonar analysis reports, lint/format checks, and ArchUnit findings (SKILL.md).
  • Boundary markers: Relies on the 'persona-plan-marshall-agent' PR review rule to validate suggestions against plan intent before processing.
  • Capability inventory: Can propose code changes (FIX), reply to PR threads, and trigger 'AskUserQuestion' for human review.
  • Sanitization: Content validation is managed by the central persona agent rather than locally in this knowledge extension.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 11, 2026, 07:13 AM
Security Audit — agent-trust-hub — ext-triage-java