Four questions before you install
If you already run auto mode, a sandbox and an allowlist, this is what ACP adds and what it leaves alone.
Why not auto mode and a sandbox?
Keep both. A sandbox bounds where the agent’s code can run. Auto mode and allowlists are settings inside one harness. Neither looks at an action the agent takes with your own credentials: a push with your git login, a DROP TABLE against a database it can reach, a curl that sends a file out.
ACP sits at the tool call itself. One set of rules covers every harness you run, the same call gets the same answer in each, and there is one log to read afterwards. Rules are set per tier, which is the kind of run: interactive, subagent, background or API. For runs with nobody watching, set the background and subagent tiers to Never on rm, git push, curl and your database clients, keep interactive sessions on Ask first, and keep production credentials out of the agent’s environment. The overnight recipe has the rules.
Does it cover my other harness?
Fourteen harnesses install with the one command (thirteen on native Windows). What differs is where you approve. Only Claude Code shows a held call in the terminal. On the other harnesses the call is blocked, you approve it on the console or from the email, then ask the agent to retry. One exception today: DeepSeek Harness reports the block to the agent but does not enforce it yet.
The safety floor is the set of built-in holds that are on with no rule written: a force-push, destructive SQL through a database client, or curl | sh waits for a person. It reads shell commands (Bash and PowerShell). File edits and MCP tools do not reach the floor; they are recorded and take your rules by tool name. What each harness lets ACP intercept is in the coverage table.
What leaves my machine?
In a connected install the hook sends each call to ACP to decide and record: the tool name and the tool’s input before the call runs (for a shell command, the command text), and up to 200 KB of the tool’s output afterwards. The stored record is the metadata: tool, who ran it, the decision and the reason, latency, tokens and cost. By default the command’s text and output are not stored; a workspace admin can turn on short previews. A held call keeps up to 500 characters of its arguments so the approver can read what they are approving. A credential that appears in a command’s arguments is redacted before the record is written.
With --local nothing leaves. Rules live in ~/.acp/policy.json and the log in ~/.acp/audit.jsonl. The privacy page lists everything that is stored.
What if ACP is down?
On Claude Code, an interactive session keeps running and shows a loud UNGOVERNED notice, meaning the call ran without a check. The lapse is written to ~/.acp/lapse.log and reported on the next checked call, and the safety floor still runs on the machine. Subagents, background runs and API sessions stay blocked until ACP answers, so set the rules for those tiers with that in mind.
Other harnesses use the shared hook, which lets calls through when ACP is unreachable unless you set ACP_FAIL_MODE=closed or ~/.acp/failmode on that machine. --local has no server to lose.
Install
To install, follow Install into your agent. It is free for up to 5 agents; pricing has the rest.
Running coding agents? ACP records each tool call its hook sees and makes the dangerous shell commands (a force-push, destructive SQL, curl | sh) wait for you. One command, free up to 5 agents:
curl -fsSL https://agenticcontrolplane.com/install.sh | bash
irm https://agenticcontrolplane.com/install.ps1 | iex