foldrun@19.0.1
Agents are just folders. Write them, check them, run them — on your machine.
AI security review
scanned · by lpm-firewall-aiThe reviewed entry point dispatches user-selected CLI commands, and the flagged process execution supports local git inspection and optional Playwright discovery. The credential references are for model runs; I found no evidence of automatic unrelated collection or a concealed attack.
Decision evidence
public snapshot- The launcher imports the CLI command module and invokes the requested command; the network and model work is command-driven.