Provides tools for managing TP-Link routers (Archer/AX series), including listing connected clients, router status, WiFi information, DHCP leases, ethernet port status, and rebooting the router via the router's web API.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@tplink-mcp list all connected clients"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
next time imma flash my own openwrt so i dont have to do this shit
A CLI + MCP tool for TP-Link Archer/AX routers, so you can manage the router from the terminal instead of the web UI. Reverse-engineered from the site's JS.
With uv, install it as a global tool - puts tplink and
tplink-mcp on your PATH:
uv tool install '.[mcp]' # from a local checkout
# uv tool install 'tplinkcli[mcp]' # once published to PyPIThen put your credentials in ~/.config/tplink/.env:
ROUTER_IP=192.168.0.1
ROUTER_USERNAME=admin
ROUTER_PASSWORD=your-router-passwordCredentials are also read from environment variables, a project-local .env, or a file
pointed to by $TPLINK_ENV (precedence, low → high: user config → project .env →
$TPLINK_ENV → env vars → --host/--password flags).
Related MCP server: home_router
The router only allows one admin session at a time, so each tplink <command> logs
in, runs, and logs out, releasing the session for the web UI. For several commands in a
row, use the interactive session instead:
tplink # interactive session:
tplink> clients
tplink> raw status?form=internet --op read # no quoting needed in here
tplink> exit # logs outOne-shot commands:
tplink clients # connected devices with IP + MAC
tplink status # mode, WAN IP, uptime, CPU/mem, SSIDs
tplink wifi # SSID + on/off per band (passwords hidden; --reveal-secrets)
tplink stats # per-client wireless signal (dBm), PHY rate, band
tplink ports # physical ethernet port link status
tplink dhcp # DHCP leases
tplink watch clients|leases # live poll: print join/leave/IP-change diffs (Ctrl-C to stop)
tplink dhcp-config # DHCP server pool / lease time / DNS
tplink reservations # DHCP address reservations
tplink syslog --level ERROR --limit 50 # system log (DHCP/SAE/DFS events)
tplink radio 5g # radio settings for a band (channel, width, security)
tplink wifi-adv 5g # advanced: WMM, beacon/DTIM/RTS, tx power, OFDMA, TWT, band steering
tplink firmware # firmware/hardware version + cloud update check
tplink wan # WAN/internet status (JSON)
tplink ipv6 # IPv6 WAN + LAN status
tplink session # session age / login count
tplink dump -o snap.json # full-state snapshot (config-drift diffing)
tplink reboot # reboot (asks to confirm; --yes to skip)
tplink raw 'status?form=client_status' --op load --json # call any endpoint directlyQuote the raw path in your shell (the ? is a glob) — not needed inside the interactive
session. Add --json to most commands for machine-readable output.
Exposes the router to an AI agent — 24 tools including get_syslog, get_client_stats,
get_dhcp_settings/set_dhcp_settings, list_reservations/add_reservation,
get_wifi_radio/set_wifi_channel/set_wifi_security, get_wps/set_wps,
get_guest_network/set_guest_network, session_info, list_endpoints, and raw_request.
Wi-Fi passwords are redacted by default (reveal_secrets=true to opt in); radio-restarting
writes require confirm=true.
tplink-mcpExample Claude Code config (credentials via ~/.config/tplink/.env or inline env):
{
"mcpServers": {
"tplink": {
"command": "tplink-mcp",
"env": { "ROUTER_IP": "192.168.0.1", "ROUTER_PASSWORD": "..." }
}
}
}Work on the code from a checkout without installing globally:
uv sync --extra mcp # dev tools (pytest) are in the default dependency group
uv run pytest -q # crypto tests run offline (synthetic vectors only)
uv run tplink status # run against the router without a global install
uv tool install --force '.[mcp]' # reinstall the global tool after changesThis server cannot be deployed
Give your AI agents the tools to build, manage, and run automation workflows.
Connect AI assistants to authorized Windows and macOS devices for files, commands, and management.
Provides capabilities that let LLM agents perform a range of infrastructure management tasks.
Discover, compare, route, and execute machine-accessible capabilities for AI agents.
