agent-client-protocol

Automate ACP integration tasks between coding agents and editor clients.

Updated Mar 2, 2026
One-click install
npx skills add https://github.com/goddard-ai/goddard --skill agent-client-protocol
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-client-protocol
Source: https://github.com/goddard-ai/goddard/tree/main/.agents/skills/agent-client-protocol
Command: npx skills add https://github.com/goddard-ai/goddard --skill agent-client-protocol

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill provides a structured approach to integrate, review, debug, or explain the Agent Client Protocol (ACP) between coding agents and editor or IDE clients. It covers initialization, capability negotiation, session lifecycle handling, prompt and update streaming, tool calls and permission requests, filesystem or terminal bridges, MCP server wiring, session modes or config options, slash commands, transports, and ACP extension points.

Core Features & Use Cases

  • Standardized JSON-RPC surface for agent-editor communication.
  • Guidance for lifecycle events: initialize, session/new, session/prompt, session/update, session/cancel.
  • Clear rules for content blocks, tool calls, permissions, and MCP transports.
  • Practical scenarios: validating protocol implementations, debugging integration issues, reviewing ACP surfaces.

Quick Start

Initialize ACP integration testing between an agent and an editor client and verify session lifecycle methods and capability negotiation.

Frequently Asked Questions about agent-client-protocol

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How does the Agent Client Protocol handle session lifecycle and capability negotiation?

ACP standardizes agent-editor communication using a JSON-RPC surface, covering initialization, capability negotiation, session lifecycle, prompt streaming, tool calls, and MCP server wiring to ensure structured and predictable integration.

How do I integrate a coding agent with an IDE client using ACP conventions?

To integrate an agent with an IDE using ACP, enforce conventions like absolute paths and 1-based line numbers, wire MCP servers, and handle tool calls, permission requests, and filesystem or terminal bridges through standardized JSON-RPC methods.

What is the best way to debug JSON-RPC communication issues between an agent and an editor?

Debug agent-editor JSON-RPC issues by validating protocol implementations, reviewing ACP surfaces, checking session lifecycle methods, and verifying content blocks, tool calls, and MCP transports against ACP conventions.

Does the Agent Client Protocol support slash commands and terminal bridges?

Yes, ACP supports slash commands, session modes, and terminal bridges, allowing coding agents to execute filesystem or terminal operations and request permissions through a standardized JSON-RPC interface.

Why is capability negotiation important when initializing an ACP session?

Capability negotiation during ACP initialization drives behavior by ensuring the agent and editor client agree on supported features, content preferences, and available transports before processing prompts or tool calls.

What are the limitations when using ACP for MCP server wiring and prompt streaming?

ACP limitations center on its strict conventions: absolute paths and 1-based line numbers are enforced, and all MCP transports, content blocks, and permission requests must adhere to negotiated capabilities established during initialization.