"Understanding Boss Mode" matching MCP servers:

Enables AI agents to interact with Fortinet FortiManager through a minimal MCP interface using sandboxed JavaScript for querying the API spec and executing JSON-RPC calls.

Enables secure, governed interaction with DMTF Redfish-compliant BMCs (iDRAC, iLO, etc.) via MCP, with read-only default and write operations gated behind explicit write-mode elevation and dry-run confirmations.

Enables AI agents to query Oracle WebLogic Server runtime state, health, and JVM metrics through the REST management API, and to start, shut down, or force-shut-down servers via Node Manager, all under a least-privilege account with allowlisting, read-only mode, and JSON audit logging for every call.

Enables an agent to administer a Linux host (with first-class Arch support) over SSH through 109 typed tools covering system state, systemd, journal, packages, network, disks, hardware, boot, sessions, users and security, all returning compact JSON. Reads are the default, while mutations require write mode plus a single-use confirmation token, with optional sudo for root-level operations.

Enables querying and managing Apache APISIX gateway configurations (routes, services, upstreams, consumers, plugins) through natural language, with read-only mode and write operations requiring confirmation.

Lets Claude or any MCP client inspect a TP-Link Deco mesh locally — nodes, backhaul, clients, WAN, firmware — and carefully change settings with graded, read-back-verified writes, secrets redacted by default and an optional read-only mode. No cloud or app involved.

Enables AI assistants to build and configure Cisco Packet Tracer topologies, run device and connectivity commands, and inspect simulation-mode packet explanations through natural language.

Enables AI clients to perform controlled diagnostics and configuration of Keenetic and Netcraze routers over LAN RCI or remote KeenDNS HTTPS, with read-only mode, guarded writes, and centralized secret masking.

Enables AI agents to control a locally running Bettbox (mihomo-based Clash client) without any client patches, switching proxy nodes, setting routing mode or TUN, testing latency, listing live connections, and adding, removing, or updating subscriptions. All capabilities ride on the client's existing surfaces: the mihomo REST API, its control socket, and its config/profile files.

Drives a running Cisco Packet Tracer over its native IPC protocol (PTMP): build and cable topologies, configure IOS, test with ping and simulation mode, and reach the whole IPC API through 72 tools.

Simulates a firewall MCP server that exposes deterministic device information, security rules, and traffic logs. In spindl mode, it spools large responses server-side so clients can query, aggregate, and explore distinct values without filling their context window.

Enables AI CLI agents to operate an aaPanel (btPanel fork) hosting panel through its signature-authenticated HTTP API, exposing 57 described operations covering websites, databases, SSL certificates, backups, and system status plus resources for panel inventories. Ships with a read-only default mode and a separate gate over credential-exposing and file-level operations, redacting generated passwords from tool results.

Enables managing Nginx Proxy Manager resources like proxy hosts, redirection hosts, streams, access lists, certificates, users, and settings through MCP, with automatic JWT refresh and configurable transport modes.

A read-only MCP server for the Cisco Meraki Dashboard API that exposes 50 GET endpoints for understanding network topology from organization down to individual devices and ports.

Enables MCP clients like Claude Code, Claude Desktop, or Cursor to read and manage Remnawave 3.x VPN panel resources — users, nodes, hosts, config profiles, squads, subscription templates, billing, and HWID devices — through the panel's REST API, with contract-driven tool schemas, numeric user IDs, multi-panel config lookup, and an optional readonly mode.

MCP server for Porkbun domains and DNS, enabling read-only and mutating operations with safety features like dry-run and write mode toggle.

Enables deploying apps, managing servers, domains/DNS, VPS lifecycle, and account balance on Impreza Host via AI in VS Code.

Enables AI assistants like Cursor, Claude, and Codex to interact with Krutrim Cloud, providing discovery and controlled operations for VPCs, compute, storage, networking, Kubernetes, KPods, and IAM, with safety controls such as confirmation and read-only mode.

Open-source MCP servers that give AI agents structured access to networking standards, device databases, and security data.

It enables topology planning, validation, auto-fix, JavaScript/CLI generation, and live deployment into a running Cisco Packet Tracer through a local HTTP bridge, with a file-mailbox fallback.

Enables reviewing a VPS over SSH (connectivity checks, service and disk status, recent service logs) and running restart, deploy, or rollback actions restricted to commands declared in a local configuration. Actions require explicit confirmation, are rate-limited and audited, and can be disabled entirely with a read-only mode.

Lets AI agents control a Keenetic router over the local network through the Model Context Protocol, managing devices, Wi-Fi, port forwarding, KeenDNS, DNS, DHCP, firewall, mesh, storage, logs and diagnostics with a safe, read-back-verified change model. An opt-in SSH mode additionally enables Entware package management and config file editing on the router.

MCP server for running infrastructure diagnostics from 6 global regions. It provides tools like SSL checks, DNS lookups, ping, whois, port checks, traceroute, latency tests, and more, with a free demo mode (10 calls/day) and no API key required.

Enables debugging of bunny.net CDN pull zones — inspecting edge rules, cache purges, statistics and origin error logs — and managing bunny Shield WAF settings, custom rules, rate limits, bot detection and WAF event logs, with credential redaction and before/after change diffs. A read-only mode can restrict it to inspection tools.

Enables AI assistants to control a local Clash Verge Rev / mihomo core on Windows over its named-pipe controller, discovering the live pipe automatically so it works in both sidecar and service mode. Supports inspecting status, switching proxy nodes, latency testing, changing modes, viewing connections and rules, and updating providers.

Enables Claude and other MCP clients to troubleshoot and, with an optional admin account, manage Citrix NetScaler / ADC appliances over the NITRO REST API, with read-only tools by default plus a sandboxed SSH path for counters and newnslog events. It is HA-aware, supports multiple appliances and local stdio or hosted HTTP with per-user roles and write auditing.

Exposes 338 tools for the Proxmox Virtual Environment API, enabling users to inspect and control nodes, QEMU VMs, LXC containers, storage and backups, clustering and HA, users and permissions, firewall rules, disks, Ceph, ACME certificates, SDN, notifications, and resource pools. Includes lifecycle actions, snapshots, migration, metrics, task tracking, and guest-agent command execution, with an optional read-only mode that blocks all writes.

An MCP (Model Context Protocol) server that enables AI assistants to manage Exoscale cloud infrastructure. Supports both stdio and Streamable HTTP transports with read-only mode enabled by default for safety.

Enables LLM agents to manage RabbitMQ queues, exchanges, and shovels via MCP tools, with restricted mode for production environments.

Enables authenticated remote MCP clients to manage a Kali Linux environment, run modular tools in sandboxed or optional host contexts, configure networking, and monitor jobs through a live WebUI.

Enables AI agents to browse organizations, networks, devices and connected clients, read switch ports, SSIDs, RF profiles, firewall rules and VPN settings, and — when an operator permits in read-only-off mode with explicit confirmation — change them or call any other Dashboard API endpoint. Credentials arrive per request as headers, nothing is stored, and every change is refused by default.

Enables AI agents to drive wireless penetration testing end-to-end through MCP, including monitor mode, network scanning, handshake and PMKID capture, WPS attacks, deauthentication, and password cracking. It exposes all h4kfi tools to MCP-compatible clients like OpenCode and Claude Code.

MCP server for managing cPanel accounts and WHM servers via UAPI/WHM APIs. Supports domains, DNS, email, MySQL, files, cron, SSL, PHP, backups, and hosting accounts across multiple environments with read-only mode and security controls.

Enables inspecting local listening ports, process owners, Docker mappings, and local LLM signatures, plus looking up or terminating the process that owns a specific port when write access is allowed. Mutating actions require confirmation and are blocked in dry-run mode.

Exposes the UniFi Network Integration API as MCP tools, dynamically loaded from JSON manifests, with read-only mode by default.

Enables AI clients to monitor EVM-compatible RPC endpoints by checking health, block synchronization, node information, uptime, and comparisons across multiple endpoints.

This MCP server provides tools to interact with FortiManager via JSON-RPC, but is deprecated and replaced by a more efficient Code Mode architecture.

MCP server that lets an AI agent manage Proxmox VE in natural language, with policy-driven security, read-only mode, two-step confirmations, and audit logging.

Read-only MCP server that surfaces UniFi network data from a UCG-Fiber console to Claude, enabling monitoring of hosts, sites, devices, clients, ISP metrics, and device stats via read-only API tools.

Telnet-based MCP server for H3C network device CLI, enabling connection, command execution, and device mode detection via natural language.