opencode-acp

What This Skill Does

The opencode-acp skill enables your OpenClaw agent to interface directly with the OpenCode environment using the Agent Client Protocol (ACP). This integration transforms your agent into a powerful pair-programmer capable of executing complex code modifications, refactoring existing architectures, and implementing new features autonomously. By abstracting the complexities of JSON-RPC communication and session lifecycle management, this skill allows the agent to focus on high-level software engineering tasks while the ACP client handles the protocol overhead.

Installation

To integrate this capability into your OpenClaw ecosystem, execute the following command in your terminal: clawhub install openclaw/skills/skills/ailois/opencode-acp Ensure you have Python 3.9 or higher installed on your system, as the skill performs automated environment setup for its dependencies. The skill will manage its own virtual environment to prevent dependency conflicts.

Use Cases

• Automated Refactoring: Easily migrate legacy functions to modern patterns without manual copy-pasting.
• Feature Implementation: Describe a desired user flow or API endpoint, and let the agent handle the implementation logic.
• Debugging & Hotfixing: Allow the agent to read source files, identify logical errors, and propose code-based solutions for immediate verification.
• Documentation Generation: Automate the creation of JSDoc or docstrings by having the agent parse your codebase and generate meaningful documentation.

Example Prompts

1. "Use opencode-acp to refactor the database connection logic in /src/db to use a connection pool instead of individual instances."
2. "Add a new POST endpoint for user registration to our current Express server and implement basic validation logic."
3. "Analyze the current project structure and propose a set of changes to implement a modular plugin architecture."

Tips & Limitations

• Pattern adherence: Always follow the Plan → Execute → Verify loop. For large tasks, ask the agent to present a plan before executing code changes.
• Scope: Ensure you point the project path to the root directory where the OpenCode server expects the workspace to be, or the tool may fail to locate source files.
• Security: As this skill involves code execution and file modifications, ensure you review the agent's proposed plan before authorizing high-impact changes to your codebase. Use the --verbose flag for auditing agent activities during development phases.