manage-mounts

Restrict Docker container filesystem access by configuring a JSON-based mount allowlist that explicitly defines which host directories are available, ensuring secure isolation for containerized agents. This prevents unrestricted access to your local filesystem.

Yes, you can enforce read-only permissions for specific host directories to prevent non-main agents from accidentally modifying data. This is configured through the mount allowlist to maintain strict filesystem isolation in multi-agent environments.

A mount allowlist is a controlled configuration mechanism that defines explicit read-write or read-only directory access for containerized AI agents. You need it when running multi-agent environments to prevent unrestricted local filesystem access.

Yes, managing secure mounts requires interaction with underlying NanoClaw configuration files and system service management to apply directory access changes and enforce filesystem isolation for your containerized agents.

The allowlist approach requires explicit configuration for every accessible directory, meaning any unlisted host paths remain completely inaccessible. You must manage JSON configuration files and coordinate with system services to apply changes.

Add a project folder to the allowlist by updating the JSON-based mount configuration to include the specific directory path and defining its permission level as read-only, which isolates the rest of your system during agent execution.