FAQ
Remote Control is for operating Claude Code from another device. Agent Tick is for least-permission command approval. When an agent tries to run a command, Agent Tick blocks it and asks your phone yes or no. Your phone cannot drive your terminal or execute arbitrary commands.
Those tools overlap in useful ways. Agent Tick focuses on the narrow command gate: block the command the agent tried to run, ask the right phone, then tell the CLI to either pass that command or stop it.
Because hosted Agent Tick sells convenience: no server, backups, updates, auth setup, push plumbing, or uptime burden. You can self-host the same command-approval flow; many teams will simply prefer not to operate another small service.
Agent Tick is built by Self-Deprecated. We are real developers using coding agents in real work, and this is the kind of small utility we wanted for our own workflow.
No. Agent Tick is deliberately narrower than that. It helps you respond to agents from your phone so work does not stall while you are away from your desk.
No. Agent Tick blocks a command before it runs and returns a yes/no decision. Your local CLI uses yes to unblock that exact command, or no to stop it.
No. The mobile app is a pass/fail control for the blocked command. It can approve that exact command or deny it; it cannot invent a new command.
The decision returns to the CLI as a failure. The blocked command does not run, and the agent has to continue without it or stop.
No. Organization plans use simple flat tiers. Agent Tick routes pings; the pricing should reflect that.
The technically correct phrase is source-available. Agent Tick uses the Business Source License: you can inspect, modify, and self-host it, but cannot offer Agent Tick itself as a competing hosted service during the BSL period.
