Skip to main content
Glama

MCACP

Bridge any MCP client to any ACP coding agent.

MCACP is an MCP server that speaks the Agent Client Protocol on behalf of your MCP host. It lets Claude Code, Zed, VS Code, or any MCP-compatible client spawn, manage, and interact with ACP coding agents — turning a single chat window into a multi-agent control plane.

MCP Client (Claude Code, Zed, VS Code, ...)
    |
    |  MCP (stdio)
    v
  MCACP
    |
    |  ACP (stdio, per agent)
    v
 Claude Code  ·  Gemini CLI  ·  Codex  ·  Auggie  ·  Copilot  ·  ...

What can you do with this?

Run any ACP agent from any MCP client. You don't need a dedicated IDE plugin for each agent. Install Gemini from Claude Code. Run Codex from Zed. Mix and match.

Compare agents side by side. Give the same task to Claude Code ACP and Gemini CLI in parallel sessions and compare the results.

Build agent-driven pipelines. One agent writes code, another reviews it, a third runs the tests. MCACP handles the lifecycle and permission flow for all of them.

Control permissions interactively. On hosts that support MCP elicitation (Zed), permission requests pop up as interactive forms. On hosts that don't (Claude Code), they fall back to operator mode where the host agent decides.

Resume where you left off. Sessions persist to disk. Shut down, restart, load_session — the agent picks up where it stopped.

Discover and install agents from chat. Browse the ACP registry, install agents, check for upgrades — without leaving your conversation.

Related MCP server: personal-mcp

Quick start

Install

npm install -g mcacp

Or run directly:

npx mcacp

Add to your MCP client

Claude Code (~/.claude/settings.json):

{
  "mcpServers": {
    "mcacp": {
      "command": "npx",
      "args": ["-y", "mcacp"]
    }
  }
}

Zed (settings.json):

{
  "context_servers": {
    "mcacp": {
      "command": { "path": "npx", "args": ["-y", "mcacp"] }
    }
  }
}

Then from chat:

"Search for available agents"        →  registry_search
"Install the Gemini agent"           →  agent_install
"Initialize it and start a session"  →  initialize + new_session
"Ask it to refactor auth"            →  prompt_start + prompt

MCP tools

MCACP exposes 22 tools organized into five groups.

Registry

Tool

Description

list_installed_agents

List locally installed agents

registry_search

Search the ACP agent registry

agent_install

Install an agent from the registry

agent_uninstall

Remove an installed agent

agent_check_upgrades

Check for available upgrades

discover_agents

Scan editor configs (Zed, JetBrains) for agents

reload_config

Reload mcacp.json from disk

Lifecycle

Tool

Description

initialize

Spawn an agent process and perform the ACP handshake

shutdown

Gracefully shut down an agent

Sessions

Tool

Description

new_session

Create a session with working directory and permission policy

load_session

Resume a persisted session

list_sessions

List stored sessions for an agent

close_session

Close a session (preserved for later resume)

Interaction

Tool

Description

prompt_start

Send a prompt (returns immediately)

prompt

Block until events are available

prompt_events

Non-blocking poll for events

grant_permission

Resolve a pending permission request (operator mode)

request_permission

Send an MCP elicitation to the host for a permission decision

cancel

Cancel an in-progress prompt

set_mode

Switch agent operating mode

Status

Tool

Description

list_running_agents

List all agents with status and sessions

get_agent_status

Detailed status of a running agent

set_agent_status

Set status text for an agent

Permission policies

Each session can use a different policy:

Policy

Behavior

allow_all

Auto-approve all agent actions

deny_all

Block all tool use

elicit

Forward as MCP elicitations to the host; falls back to operator if unsupported

operator

Queue as events — the host agent decides via grant_permission

Operator flow

Orchestrator                   MCACP                      ACP Agent
    |                            |                            |
    |-- prompt(sessionId, ...) ->|                            |
    |                            |-- session/prompt --------->|
    |                            |<- request_permission ------|
    |<- {pendingPermission: ...} |                            |
    |                            |                            |
    |-- grant_permission ------->|                            |
    |   {toolCallId, optionId}   |-- permission outcome ----->|
    |                            |<- prompt result -----------|
    |<- {stopReason, updates}    |                            |

Configuration

MCACP looks for mcacp.json in these locations (first found wins):

  1. --config flag or MCACP_CONFIG env var

  2. .mcacp/mcacp.json in the current directory

  3. ~/.config/mcacp/mcacp.json

{
  "registries": ["https://cdn.agentclientprotocol.com/registry/v1/latest/registry.json"],
  "defaultPermissionPolicy": "elicit",
  "agent_servers": {
    "my-custom-agent": {
      "command": "node",
      "args": ["my-agent.js"]
    }
  }
}

See docs/configuration.md for the full schema.

Building from source

git clone https://github.com/Oortonaut/mcacp.git
cd mcacp
npm install
npm run build
node dist/index.js

License

Apache-2.0

Available Tools

24 tools
agent_check_upgradesB

Check all installed agents for available version upgrades.

Inputs: no named parameters.

agent_installB

Install an ACP agent from the registry.

Inputs: agentId (required), version.

agent_uninstallB

Remove a locally installed ACP agent.

Inputs: agentId (required).

cancelB

Cancel an in-progress prompt.

Inputs: sessionId (required).

close_sessionA

Close an active session. Session file preserved for future load_session.

Inputs: sessionId (required).

discover_agentsA

Scan editor configs (Zed settings.json, JetBrains acp.json) for ACP agents. Returns agents with command/args/env and source. To import, add them to agent_servers in your mcacp.json and call reload_config.

Inputs: no named parameters.

eventsA

Block until any prompted session produces events. Returns events stamped with sessionId and agentId. Supports optional Nagle-style coalescing to batch events across sessions.

Inputs: timeoutMs, nagleMs.

get_agent_statusB

Get detailed status of a running ACP agent.

Inputs: agentId (required).

grant_permissionA

Respond to a pending permission request (operator policy). The agent resumes; new events flow to the queue.

Inputs: sessionId (required), toolCallId (required), optionId (required).

initializeA

Spawn an ACP agent process and perform the initialize handshake. Must be called before creating sessions.

Inputs: agentId (required), protocolVersion, clientInfo, clientCapabilities, +1 more.

list_installed_agentsA

List all locally installed ACP agents with their id, name, version, and description.

Inputs: no named parameters.

list_running_agentsA

List all spawned ACP agent processes with status, heartbeat, and active sessions.

Inputs: no named parameters.

list_sessionsB

List stored sessions. Does not require the agent to be running.

Inputs: agentId (required).

load_sessionC

Resume a previously created session. Agent must support session loading.

Inputs: agentId (required), sessionId (required), cwd (required), mcpServers.

new_sessionA

Create a new ACP session on an initialized agent. Returns the agent's session ID.

Inputs: agentId (required), cwd (required), mcpServers, permissionPolicy.

prompt_eventsA

Non-blocking poll for prompt events. Returns all queued events (updates, permission requests, completion). May return empty array if no new events.

Inputs: sessionId (required).

prompt_polledA

Send a prompt to an active ACP session. Returns immediately with { status: "prompted" }. Use prompt_events to poll for events. Pair with prompt_events for the async pattern; use prompt_sync for the blocking pattern.

Inputs: sessionId (required), prompt (required).

prompt_syncA

Send a prompt and block until it completes, errors, or a permission_request needs operator attention. Returns all collected events at once. Returns early (without a complete event) on permission_request or timeout. For the async pattern, use prompt_polled + prompt_events instead.

Inputs: sessionId (required), prompt (required), timeoutMs, includeThoughts, +1 more.

reload_configA

Reload configuration from disk. Use after editing mcacp.json to pick up new agent_servers entries or changed settings. Returns the new config.

Inputs: no named parameters.

request_permissionB

Send an MCP elicitation to the outer host asking for a permission decision. Does not require a running agent or session.

Inputs: title (required), options.

set_agent_statusA

Set an explicit status text for a running agent.

Inputs: agentId (required), status (required).

set_modeC

Switch the operating mode of a session.

Inputs: sessionId (required), modeId (required).

shutdownB

Gracefully shut down a running ACP agent, closing all sessions.

Inputs: agentId (required).

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.6/5.0

    Scored across 24 tools

    Disambiguation5/5

    Each tool has a clearly distinct purpose, covering separate aspects like agent lifecycle, session management, prompting patterns, permissions, and configuration. No two tools overlap significantly; even similar-sounding tools like prompt_polled, prompt_sync, and prompt_events are well-differentiated by their descriptions.

    Naming Consistency4/5

    Most tool names follow a consistent verb_noun or verb_adjective_noun pattern in snake_case (e.g., agent_install, list_sessions, prompt_sync). A few outliers like 'events' (noun-only) and 'cancel' (verb-only) exist, but the overall pattern is strong and intuitive.

    Tool Count4/5

    With 24 tools, the count feels appropriate for the complex domain of ACP agent management. Each tool addresses a distinct operation, and the number is high enough to cover the required functionality without being excessive.

    Completeness4/5

    The tool set covers the core lifecycle (install, uninstall, initialize, shutdown, status), session management (new, close, load, list), multiple prompting modes (sync, async), permissions, and configuration. Minor missing operations like an explicit update tool could be added, but the set is largely comprehensive.

    Maintenance

    ActivityInactive
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      F
      maintenance
      The self-hosted MCP bridge between Claude Chat and Claude Code.
      46
      AGPL 3.0
    • F
      license
      Not graded
      quality
      B
      maintenance
      MCP server that bridges coding agents (Claude Code, Codex, Gemini CLI) via ACP for pair programming, enabling agents to consult each other as tools.
      3
      -