claude_cli_skill
A universal MCP-compatible tool/skill that empowers ANY AI Agent to natively call the local Anthropic Claude Code CLI. Ideal for code refactoring, large scale code generation, terminal execution, bash shell commands, running tests, local environment modification, and bypassing token limits.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/bighit1/use-claudecodeWhat This Skill Does
The claude_cli_skill acts as a powerful bridge between the OpenClaw AI agent and the Anthropic claude-code CLI tool. It transforms your local terminal into an autonomous coding environment, allowing OpenClaw to interact with your codebase as a human developer would. By delegating complex engineering tasks to this skill, OpenClaw can navigate directories, refactor large-scale projects, fix bugs, and execute bash commands without hitting traditional token limits or struggling with internal context windows. The skill operates by spawning persistent sessions within your specified project_path, effectively treating the CLI as a sub-agent dedicated to local file-system operations and terminal-based problem solving.
Installation
To integrate this capability into your OpenClaw environment, ensure you have the claude-code CLI installed globally on your machine. Once confirmed, install the skill via the OpenClaw plugin manager using the following command:
clawhub install openclaw/skills/skills/bighit1/use-claudecode
Use Cases
This skill is indispensable for scenarios involving complex engineering workflows. Use it when you need to:
- Architect large software applications from scratch.
- Refactor sprawling codebases that require multiple file updates across different directories.
- Debug intricate runtime errors that require testing and reproducing the issue in the local environment.
- Automate routine maintenance tasks, such as dependency updates or linting, where terminal access is mandatory.
Example Prompts
- "OpenClaw, please scan the current directory and refactor the entire authentication module in /projects/app to use OAuth2 instead of local tokens."
- "I'm getting a TypeError in my unit tests. Run the test suite in /home/user/work/my-app using Claude CLI and analyze the stack trace to suggest a fix."
- "Start a new coding project in /tmp/new_tool that implements a basic CLI utility for file renaming."
Tips & Limitations
- Persistence: Use
is_new_session: falseto maintain context across multiple messages; useis_new_session: trueonly when switching to a fundamentally different project. - Latency: Be aware that because this tool interacts with a real CLI and performs file system operations, results may take several seconds or minutes to return.
- Safety: While the tool automates
bypassPermissions, always review changes suggested by the agent before finalizing major destructive commits.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-bighit1-use-claudecode": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: file-write, file-read, code-execution, external-api