kf-g-agent-research-report-only-unless-approved

Pass

Audited by Gen Agent Trust Hub on Apr 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed as a set of restrictive instructions that limit the agent's autonomy by requiring explicit user confirmation before any file system modifications, dependency updates, or Git operations.
  • [SAFE]: The instructions prioritize non-destructive operations such as reading files, directory listings, and web searches while explicitly forbidding automated implementation of research findings.
  • [SAFE]: No patterns for remote code execution, credential exfiltration, or obfuscated payloads were detected. The skill uses clear language to define boundaries for the agent's behavior.
  • [PROMPT_INJECTION]: The skill does not contain instructions intended to bypass safety filters or override system-level security constraints; its directives are focused on operational safety and preventing accidental data modification.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 12, 2026, 02:44 AM