Filesystem

What This Skill Does

The Filesystem skill by amaofx is an advanced toolkit designed for OpenClaw AI agents to interact with, analyze, and manipulate local storage environments. It provides a robust interface for performing complex operations ranging from simple file listing and directory navigation to sophisticated batch processing and content extraction. The skill wraps native shell utilities like find, grep, du, and stat to provide the AI with a granular view of the disk structure, enabling it to manage assets, clean up redundant files, organize code repositories, and extract actionable data from vast directory trees without needing manual human intervention.

Installation

To integrate this skill into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/amaofx/filesystem Ensure that your agent has the necessary permissions to access the target directories, as the skill operates within the host system's filesystem context.

Use Cases

This skill is ideal for scenarios requiring automated disk management or data ingestion. Use it for: 1. Cleaning up build artifacts (e.g., deleting node_modules or .tmp files) across a project. 2. Auditing directory structures to identify large files or disk space hogs. 3. Searching for specific configuration strings or error logs across multiple source code files. 4. Batch renaming or organizing media and documents based on file type or creation date. 5. Generating summaries of directory contents to understand project architecture.

Example Prompts

1. "Analyze the current project directory and list the top 10 largest files so I can identify what is consuming the most disk space."
2. "Search for every instance of 'TODO' in the src directory, excluding node_modules, and output the filenames along with the line numbers."
3. "Find all image files older than 30 days in the /uploads folder and move them to a backup archive directory."

Tips & Limitations

When using batch operations like delete or move, ensure the agent has a clear scope defined to avoid unintentional data loss. Always verify the path context, as the skill defaults to the agent's current working directory. For large-scale operations on deep trees, prefer using filters (like --maxdepth) to optimize performance. Note that this skill requires direct filesystem access; it is best practice to run the agent in a containerized or restricted environment if you are working with sensitive production data.