jni-kotlin-pro

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [SAFE]: No malicious patterns or behaviors detected across the skill files. The skill provides high-quality technical guidance focused on memory safety and architectural integrity.
  • [COMMAND_EXECUTION]: The skill recommends using standard developer tools such as nm for inspecting shared library symbols and gradlew for executing unit tests. These actions are legitimate for JNI development and auditing.
  • [PROMPT_INJECTION]: The skill utilizes a governance framework that reads project-specific configuration files and scans source code for markers like // STABLE: to enforce stricter development protocols. This constitutes an indirect prompt injection surface.
  • Ingestion points: Local files docs/audit_native_jni.md, docs/engine_rules.md, and project source files.
  • Boundary markers: None; the skill relies on pattern matching without delimiters for untrusted content.
  • Capability inventory: Generates native code and recommends execution of diagnostic CLI tools (nm, gradlew).
  • Sanitization: None; logic is driven directly by the presence of markers in ingested text.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 03:34 AM
Security Audit — agent-trust-hub — jni-kotlin-pro