skills/gosukiwi/flow/flow-execute/Gen Agent Trust Hub

flow-execute

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is purely instructional and contains no executable code, shell scripts, or external dependencies. It defines a workflow for an AI agent to manage tasks via subagents.
  • [SAFE]: No network operations, data exfiltration patterns, or sensitive credential access were detected. All file references (docs/flow/plans/..., flow-shared/references/...) point to internal project documentation paths.
  • [SAFE]: The skill uses platform-specific configuration (disable-model-invocation: true) to restrict the execution environment as intended by the workflow design.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 03:32 PM
Security Audit — agent-trust-hub — flow-execute