A secure local filesystem interface that enables AI agents to read, write, and list files strictly within the designated workspace environment.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install filesystem-access-greaterpeter
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install filesystem-access-greaterpeter using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
Filesystem Access (Greaterpeter) is a foundational capability designed for AI agents to interact with local storage safely and efficiently. By providing a controlled interface for file operations, it allows agents to manage data without compromising system integrity. This skill is a core component of the broader Openclaw Skills ecosystem, ensuring that all automated file interactions are restricted to the authorized workspace directory.
The skill balances high-performance file management with strict security protocols. It is specifically engineered to prevent unauthorized directory traversal, ensuring that the AI agent remains within the logical boundaries of the project it is currently assisting with.
To enable this skill within your environment, ensure your workspace directory is correctly configured. Use the following command to initialize the environment where Openclaw Skills operate:
# Navigate to your workspace
cd /path/to/your/openclaw-workspace
# The skill is typically enabled via the agent configuration file
# Ensure the filesystem-access-greaterpeter skill is listed in your skill manifest
The skill follows a strict relative-path taxonomy to organize data. Below is the structure of how file operations are managed:
| Attribute | Description |
|---|---|
| Root Directory | Defined by the Openclaw workspace environment variable. |
| Pathing | Strictly relative paths (e.g., ./logs/output.txt). |
| File Metadata | Standard Unix-style permissions are respected for all generated files. |
| Concurrency | Handles atomic write operations to prevent file corruption during multi-agent tasks. |
../ to access files outside the authorized workspace.Loading
A professional-grade calendar integration for managing ICS-backed reminders and event feeds through a dedicated worker API.

An automated research assistant that fetches, filters, and recommends high-value AI agent and reinforcement learning papers from OpenReview and arXiv.

An AI-driven translation skill that utilizes parallel subagents to convert entire books into multiple languages and formats.

A comprehensive security auditing tool for OpenClaw environments that identifies configuration risks and provides interactive remediation.

An AI-powered image generation skill that transforms text prompts into high-quality visual assets using the Coze SDK.

An automated AI agent skill designed to fetch trending tech news and transform it into high-engagement Xiaohongshu posts.








































