skills/frizzlefur/flowkit/flow/Gen Agent Trust Hub

flow

Pass

Audited by Gen Agent Trust Hub on Sep 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill enforces security best practices by utilizing the platform's 'Plan Mode' (EnterPlanMode), which creates a read-only sandbox for the planning phase. This prevents the agent from making unreviewed modifications to the codebase until the user approves the generated task plan.
  • [COMMAND_EXECUTION]: The skill uses local shell commands via the Bash tool for environment management and session orchestration. Specifically, it checks for the presence of a tmux session ($TMUX) and uses tmux commands (list-panes, kill-pane) to manage and clean up sub-agent windows. These operations are limited to session management and do not involve sensitive system modifications.
  • [SAFE]: The skill orchestrates complex tasks by delegating to other locally installed skills such as /prompt, /multi-agent, and planning-with-files. It does not attempt to download or execute code from untrusted remote sources.
  • [SAFE]: External references are limited to the author's own GitHub repository (frizzlefur/flowkit) for documentation purposes, which aligns with the provided vendor context. No unauthorized data exfiltration or suspicious network activity was identified.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 10, 2026, 10:02 AM
Security Audit — agent-trust-hub — flow