This record as markdown: /tools/openclaw/acp.md
What acp does on OpenClaw
AI agents invoke acp to trigger actions in OpenClaw. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call: builds kicked off, notifications sent, workflows started.
Why acp is rated High
This tool orchestrates the execution of coding agents, which are capable of running arbitrary code. While the description is sparse, the combination of 'run' and 'agents' (especially 'coding agents') clearly places it in the Execute category. The severity is high because a compromised agent or malicious prompt could lead to uncontrolled code execution with potentially broad blast radius.
From the tool's definition Run and manage ACP-backed coding agents: the verb 'run' combined with 'coding agents' indicates the tool executes code or triggers external operations whose effects depend on arguments.
Attacks that exploit this kind of access
The rule that runs acp safely
PolicyLayer is an MCP gateway: it sits between your AI agents and OpenClaw, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For acp, this is the rule to start with:
acp stays usable, but rate-capped: a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.
The button opens the PolicyLayer dashboard: create your workspace, connect OpenClaw, apply this rule, and every acp call is checked against it from then on.
Questions about acp
Run and manage ACP-backed coding agents. It is categorised as a Execute tool in the OpenClaw MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the OpenClaw MCP server in PolicyLayer and add a rule for acp: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches OpenClaw. Nothing to install.
acp is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the acp rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for acp. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
acp is provided by the OpenClaw MCP server (openclaw). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on OpenClaw, and thousands of servers like it.
Across the catalogue