coala-client
How to use the coala-client CLI for chat with LLMs, MCP servers, and skills. Use when the user asks how to use coala, run coala chat, add MCP servers, import CWL toolsets, list or call MCP tools, import or load skills, or use the sandbox run_command tool.
Why use this skill?
Master the Coala Client CLI to manage LLMs, connect MCP servers, import CWL toolsets, and run modular AI skills efficiently within the OpenClaw ecosystem.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/hubentu/skills-4What This Skill Does
The coala-client skill is the official command-line interface for the Coala ecosystem, designed to bridge the gap between Large Language Models (LLMs) and local or remote tools. It functions as a powerful orchestrator that enables your AI agents to interact with the Model Context Protocol (MCP), allowing them to consume toolsets, execute complex workflows, and access real-time data from disparate sources. By providing a unified interface for OpenAI, Gemini, Ollama, and custom providers, coala-client simplifies the integration of external intelligence into your local environment. It manages environment configurations, MCP server connections, and persistent skill storage, acting as the primary control plane for your agent-assisted development.
Installation
To integrate this skill into your environment, use the OpenClaw management utility. Execute the following command in your terminal: clawhub install openclaw/skills/skills/hubentu/skills-4. Once installed, perform the initial setup by running coala init. This command establishes the necessary configuration directory at ~/.config/coala/ and creates the mcp_servers.json file and the env file where your API keys and provider-specific configurations will reside. Ensure your required API keys (e.g., OPENAI_API_KEY) are exported in your environment variables to allow the client to authenticate with external LLM providers.
Use Cases
Use coala-client when you need to extend an LLM's capabilities with specific domain tools. This includes importing CWL (Common Workflow Language) toolsets to run bioinformatics pipelines, integrating external data APIs via MCP, or managing modular AI skills for specialized tasks. It is ideal for developers who need a persistent, local chat interface that understands their specific infrastructure. Additionally, it serves as an excellent sandbox for running shell commands safely while maintaining LLM-controlled context.
Example Prompts
- "coala mcp-import my-tools https://example.com/tools.cwl to add these utilities to my active session."
- "How can I list all currently connected MCP servers and their available tool schemas using the coala CLI?"
- "Load the data-analysis skill using /skill and help me process the CSV file in the current working directory."
Tips & Limitations
To maintain optimal performance, keep your mcp_servers.json file organized and remove unused entries to prevent connection latency. While the tool is highly flexible, remember that it relies on the availability of the underlying MCP servers; if a server fails to start, the client may be unable to execute related tools. Always verify your env file for correct key formatting, and use the --no-mcp flag if you want to perform a clean chat session without loading external tool dependencies. When using the sandbox, ensure the commands being executed align with your local system policies to avoid unintended modifications to your file structure.
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-hubentu-skills-4": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: network-access, file-write, file-read, external-api, code-execution