An intelligent AI programming assistant powered by the MiniMax M2.5 model for seamless file operations, command execution, and automated code development.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install mini-agent
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install mini-agent using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Mini-Agent?
Mini-Agent is a sophisticated intelligent assistant framework specifically architected for the Openclaw Skills ecosystem and powered by the MiniMax M2.5 large language model. It functions as a bridge between natural language instructions and complex system actions, enabling developers to automate file manipulations, execute shell commands, and manage background processes with high precision.
Designed to integrate directly into the OpenClaw Agent system, this skill provides a robust environment for autonomous task completion. Whether you are refactoring code, managing system services, or organizing technical documentation, Mini-Agent utilizes its deep understanding of programming logic to act as a reliable co-developer within your workspace.
Mini-Agent Use Cases
- Automated file creation and precise string-replacement editing.
- Executing and monitoring long-running background processes via bash.
- Rapid code generation, bug fixing, and optimization across various programming languages.
- Querying usage instructions for other installed Openclaw Skills within the agent environment.
- Managing technical notes and session-specific knowledge for future reference.
How Mini-Agent Works
- The agent receives a natural language command from the user via the dashboard or API.
- The MiniMax M2.5 LLM analyzes the request to formulate a logical execution plan.
- The skill invokes specific internal tools such as read_file, write_file, or bash based on the task requirements.
- It processes the results of these tool executions to determine if further actions are necessary.
- The agent iterates through this cycle until the objective is achieved and then returns a final response.
Mini-Agent Setup
To install this skill, ensure you have the uv tool installed, then run the following command:
uv tool install git+https://github.com/MiniMax-AI/Mini-Agent.git
After installation, configure your credentials in the configuration file located at ~/.mini-agent/config/config.yaml:
api_key: "your-api-key"
api_base: "https://api.minimaxi.com"
model: "MiniMax-M2.5"
Mini-Agent Data Schema & Taxonomy
Mini-Agent organizes its operational data using a structured filesystem approach within the user home directory. The following table describes the primary data components:
| Path | Description |
|---|---|
~/.mini-agent/config/config.yaml |
Core configuration file containing API keys and model settings. |
~/.mini-agent/log/ |
Directory containing timestamped execution logs (e.g., agent_run_YYYYMMDD.log). |
/home/pi/.openclaw/agents/xiaoma |
Default workspace for file operations and relative path resolution. |
Logs capture the complete lifecycle of each request, including raw LLM messages, tool call definitions, and execution results.
Mini-Agent Advanced Features
- Support for large file handling using offset and limit parameters in read operations.
- Background process management allowing the agent to start, monitor, and kill persistent services.
- Precise file editing using exact string replacement to prevent unintended code side effects.
- Integrated skill discovery to fetch metadata and instructions from other Openclaw Skills.
- Persistent session memory that allows the agent to recall previous operations and results within a single conversation.
SKILL.md
Loading
Related Openclaw Skills
Claude Code + MiniMax-M2.5 Integration
A professional configuration guide to empower the Claude Code CLI with the high-performance MiniMax-M2.5 language model for advanced AI programming.
Million Finney Homepage Pixel Guide
A comprehensive educational guide for agents to understand the technical process of claiming pixels and generating art on the Million Finney Homepage.
Asana
A dependency-free Node.js CLI skill for managing Asana tasks, projects, and custom fields using Personal Access Tokens.
Asana
A dependency-free Node.js CLI skill for managing Asana tasks, projects, briefs, and timelines via Personal Access Tokens.
MiniMax MCP Skill
An integration for the MiniMax Coding Plan API providing real-time web search capabilities and advanced visual language modeling for AI agents.
AIGo Hotel Search
An intelligent hotel search skill that enables AI agents to query real-time lodging data, pricing, and amenities via the Model Context Protocol.
Featured*
