Loading connector details…
Loading connector details…
Choose a unique username to continue using AgentHotspot
by lucamarien • Uncategorized
A secure MCP server for managing OPNsense firewalls via AI assistants and MCP-compatible tools.
Securely query and manage OPNsense firewall configurations and status.
Automate firewall rule changes with safety via savepoints and rollback.
Integrate OPNsense management into AI workflows using MCP protocol.
This MCP server provides a secure, API-only interface to manage OPNsense firewall configurations and services through AI assistants like Claude Code and Cursor. It supports 62 tools across 10 domains including system, firewall, network, DNS, DHCP, VPN, HAProxy, services, diagnostics, and security. The server emphasizes security with read-only default mode, savepoint/rollback for changes, endpoint blocklisting, and no direct shell or config file access. It supports OPNsense 24.7 and newer versions with automatic API endpoint detection.
List all available OPNsense modules with descriptions and tool counts. In read-only mode, only read-only tools are loadable — counts reflect that. Call this first, then describe_module(), then load_module() or load_tool().
List all tools in a module with names, descriptions, and read-only status. In read-only mode, mutating tools are shown but marked as unavailable. Args: module: Module name (e.g. 'cron', 'firewall'). Get names from list_modules().
Load a single tool into the active session so it can be called. Mutating tools cannot be loaded when the server is in read-only mode. Args: tool_name: Exact tool name (e.g. 'cron_search_jobs'). Get names from describe_module().
Load all tools in a module into the active session at once. In read-only mode, only read-only tools in the module are loaded. Args: module: Module name (e.g. 'cron', 'firewall'). Get names from list_modules().
Create a system tunable. data_json: JSON with tunable fields (tunable, value, descr).
Scores are informational only and provided “as is” without warranty. AgentHotspot assumes no liability for actions taken based on these ratings.