manage-mounts

Manage host directory mount points in the NanoClaw mount-allowlist.json file.

Updated Jun 27, 2026
One-click install
npx skills add https://github.com/VincentChong123/nanoclaw-v2-agy-api-proxy --skill manage-mounts-vincentchong123
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: manage-mounts
Source: https://github.com/VincentChong123/nanoclaw-v2-agy-api-proxy/tree/main/.claude/skills/manage-mounts
Command: npx skills add https://github.com/VincentChong123/nanoclaw-v2-agy-api-proxy --skill manage-mounts-vincentchong123

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires tsx.

What problem does it solve?

This Skill addresses the security risk of granting AI agents unrestricted access to your host filesystem by providing a structured way to manage and audit directory mount points.

Core Features & Use Cases

  • Granular Access Control: Explicitly define which host directories are accessible to agent containers.
  • Read-Only Enforcement: Configure specific mount points as read-only for non-main agents to prevent accidental data modification.
  • Use Case: If you want an agent to analyze your local project logs without risking the integrity of your source code, use this Skill to mount the logs directory as read-only.

Quick Start

Ask the assistant to add the documents folder to the mount allowlist with read-only permissions.

Frequently Asked Questions about manage-mounts

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I restrict Docker container file access to specific host directories?

To restrict Docker container file access, you can configure host directory mount points using an allowlist. This validates path existence and enforces read-only constraints for non-main agents to ensure secure isolation.

What is the best way to mount a local project directory as read-only for an AI agent?

Mounting a local directory as read-only for AI agents is managed by adding the path to an allowlist file with read-only permissions. This prevents accidental data modification while allowing the agent to analyze files.

How do I audit filesystem access paths for NanoClaw agent containers?

You can audit filesystem access paths for NanoClaw agent containers by reviewing the mount-allowlist.json configuration file. This allows you to verify which host directories are explicitly accessible and check their read-only enforcement status.

Can I dynamically add or remove host directory mounts without restarting the agent container?

Yes, you can add or remove host directory mounts dynamically by updating the mount allowlist configuration. The Skill facilitates modifying these filesystem access paths within the designated JSON configuration file.

Does NanoClaw enforce read-only mounts for all agent containers?

NanoClaw enforces read-only mounts specifically for non-main agents to prevent unauthorized data modification. Main agents can be granted standard read-write access to the explicitly defined host directory mount points.

Why should I use a mount allowlist instead of granting full host filesystem access to containers?

Using a mount allowlist mitigates the security risk of granting AI agents unrestricted host filesystem access. It provides granular access control by explicitly defining accessible directories and enforcing read-only constraints where necessary.