agentic-control-plane

Acp Install — Development skill for Claude Code

Development community

One-command install for Agentic Control Plane — see, price, and control every tool call your coding agent makes.

How to install Acp Install

This entry records only its repository, not the path inside it, so there is no exact command to give. Open agentic-control-plane/acp-install and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Acp Install does

One-command install for Agentic Control Plane — see, price, and control every tool call your coding agent makes. Claude Code, Cursor, OpenAI Codex, OpenClaw, MCP clients.

Alternatives in Development

OpenAI Codex CLI — (Rust implementation) 67.8k ★
Agentic Stack — One brain, many harnesses 2.2k ★
Zeron — A native control plane for Claude Code, Codex, Cursor, Devin and other coding agents 1.4k ★

README

acp-install

**Control the tool call, control the agent.** See and stop what your coding agent actually does — every command, edit, and fetch — with one control layer that works the same across Claude Code, Cursor, and Codex. Free for individuals — connected console by default, or fully on-device with `--local`.

Real terminal recording: one command installs ACP local mode with no signup; a coding agent's force-push to main is blocked by the safety floor, a network call has to ask, normal work runs and is logged — and the same policy stops Codex too. tail ~/.acp/audit.jsonl shows every call and its decision.

Real recording, nothing mocked — how it's made / re-record it yourself. 60 seconds: install → agent blocked → asked → allowed → the audit log.

curl -sf https://agenticcontrolplane.com/install.sh | bash

Prefer fully on-device — no account, nothing leaves your machine? Add the flag:

curl -sf https://agenticcontrolplane.com/install.sh | bash -s -- --local

One command. It detects whichever agents you run and puts the same guardrails in front of all of them:

A safety floor nothing can cross — rm -rf /, mkfs, dd to a disk, a fork bomb, a force-push to main are blocked regardless of your policy (and regardless of how the command is spelled).
Your rules, one place — allow / ask / deny per tool, applied identically to every agent: the shared console (connected), or ~/.acp/policy.json (--local).
A log of what actually happened — the live activity feed in the console, or in --local mode tail -f ~/.acp/audit.jsonl, fully on-device.

The same policy that stops Claude Code stops Codex. You configure control once, not once per vendor. The default install connects a workspace — free for individuals — with the cost X-ray and a shared console across everyone's a