A Go-based Model Context Protocol (MCP) skill that allows AI agents to securely perform local file system operations such as directory listing, reading, and writing.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install system-file-handler
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 system-file-handler 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 Go FS MCP Skill?
The Go FS MCP Skill is a high-performance filesystem integration tool designed specifically for AI agents running in the OpenClaw ecosystem. Built in Go, it utilizes the Model Context Protocol (MCP) over stdio to separate arguments validation from direct I/O execution. This decoupling ensures a highly reliable and secure environment for AI-assisted file operations.
By implementing this skill, developers can grant their AI agents structured, safe access to local files and directories without manual intervention. The skill translates natural language requests into concrete file actions using a dedicated validator binary (go-fs-mcp-skill) and an execution server (go-fs-mcp-server), streamlining automation inside Openclaw Skills environments.
Go FS MCP Skill Use Cases
- AI-Driven Codebase Exploration: Allow AI agents to list directories and read files to understand project layouts and analyze source code.
- Automated Document Generation: Enable agents to write outputs, update documentation, or generate configuration files dynamically.
- File Reorganization: Request the AI agent to move, rename, or delete obsolete files securely within defined paths.
- Multi-Environment File Management: Seamlessly run file operations across Windows and WSL environments using native Go binaries.
How Go FS MCP Skill Works
- Triggering: The user interacts with the AI agent using natural language triggers (e.g., 'list directory' or 'read file').
- Validation: The OpenClaw conversation engine routes the request to the
go-fs-mcp-skillbinary, which validates the JSON arguments and encapsulates them. - Relaying: The skill forwards the validated request via MCP stdio to the
go-fs-mcp-serverbackend. - Execution: The
go-fs-mcp-serverperforms the actual I/O operations (such as listing, reading, or writing) against the local file system. - Response: The execution results are wrapped and returned to the OpenClaw chat interface in a standardized format.
Go FS MCP Skill Setup
To get started, compile the required binaries from the source repository and register them in your environment.
1. Compile the Binaries
Clone the repository and compile the CLI skill validator and the MCP server:
git clone https://github.com/go-fs-mcp/go-fs-mcp.git
cd go-fs-mcp
go build -o go-fs-mcp-server/go-fs-mcp-server ./go-fs-mcp-server/cmd/server
go build -o go-fs-mcp-skill/go-fs-mcp-skill ./go-fs-mcp-skill/cmd/skill
Alternatively, install them directly via Go:
go install github.com/go-fs-mcp/go-fs-mcp-server/cmd/server@latest
go install github.com/go-fs-mcp/go-fs-mcp-skill/cmd/skill@latest
2. Configure skill.json
Edit skill.json in your skill directory to specify the absolute path of go-fs-mcp-server.
For WSL:
"config": {
"mcp_command": "/home/user/.openclaw/skills/go-fs-mcp-skill/go-fs-mcp-server",
"mcp_args": [],
"timeout": 10
}
For Windows:
"config": {
"mcp_command": "e:\\\\github\\\\go-fs-mcp\\\\go-fs-mcp-server\\\\go-fs-mcp-server.exe",
"mcp_args": [],
"timeout": 10
}
3. Enable in OpenClaw
Enable the skill by updating your main OpenClaw configuration file (~/.openclaw/openclaw.json):
{
"skills": {
"entries": {
"go-fs-mcp-skill": { "enabled": true }
}
}
}
Restart the gateway to apply changes:
openclaw gateway restart
Go FS MCP Skill Data Schema & Taxonomy
The skill exposes 6 core tools mapped directly to Model Context Protocol actions. Below is the metadata structure for tool inputs:
| Tool Name | Required Parameters | Purpose |
|---|---|---|
list_directory |
path |
Retrieve file names and subdirectories inside the specified directory path. |
read_file |
path |
Read UTF-8 encoded text content from the specified file. |
write_file |
path, content |
Write or completely overwrite file contents. |
create_directory |
path |
Recursively create a directory structure. |
delete_file |
path |
Safely remove files or empty directories. |
move_file |
source, destination |
Rename or transfer files between paths. |
CLI Response Schema
When invoking the command line utility, the output format is returned in JSON format:
{
"success": true,
"message": "MCP 工具 list_directory 执行成功",
"data": {
"toolName": "list_directory",
"result": "..."
}
}
Go FS MCP Skill Advanced Features
- Decoupled Architecture: Leverages separate validation (
go-fs-mcp-skill) and execution (go-fs-mcp-server) binaries to minimize risk and manage runtime errors safely. - Multi-Platform Support: Ready out-of-the-box for WSL and Windows paths with configurable environment templates.
- Direct MCP Integration: Fully standardized with standard Model Context Protocol (stdio) communication channels.
- Robust JSON CLI Interface: Supports testability and automation scripts via explicit JSON CLI invocations before registering with Openclaw Skills.
SKILL.md
Loading
Related Openclaw Skills
Chukonu Patent Search
An MCP-powered patent search skill for retrieving and analyzing Chinese (CNIPA) and European (EPO) patents directly within your AI workflow.
I Have a Dream - Rhetoric & Civil Rights
An interactive AI agent skill designed to analyze, teach, and guide users through the rhetorical brilliance, moral philosophy, and historical context of Dr. Martin Luther King Jr.'s historic speech.
Imagine If... Creating a Future for Us All
An AI-powered knowledge agent skill based on Sir Ken Robinson's final book, designed to help users understand and apply the principles of creativity, imagination, and educational transformation.
In Sheep's Clothing: Dealing with Manipulative People
An AI-powered behavioral analysis and self-defense tool based on Dr. George Simon's framework for identifying and disarming covert-aggressive manipulation.
Overcoming Male Depression (I Don't Want to Talk About It)
An AI coding agent skill designed to help users explore, understand, and apply Terrence Real's clinical framework for identifying and overcoming covert male depression.
DeepRead Resume Parser
A highly precise visual and semantic resume parsing tool that converts complex CV layouts into structured, validation-ready JSON data.
METADATA
Featured*
