mcp-server/README.md
An MCP server for working with the Meshtastic firmware repo and connected devices. Lets Claude Code / Claude Desktop:
pio device monitor (with board-specific exception decoders)meshtastic Python API: owner name, config (LocalConfig + ModuleConfig), channels, messaging, reboot/shutdown/factory-resetesptool, nrfutil, picotool directly when PlatformIO doesn't cover the operationPlatformIO first. Its pio run -t upload knows the correct protocol, offsets, and post-build chain for every variant in variants/. Direct vendor-tool wrappers (esptool_*, nrfutil_*, picotool_*) exist as escape hatches for operations pio doesn't cover (blank-chip erase, DFU .zip packages, BOOTSEL-mode inspection).
pio on $PATH or at ~/.platformio/penv/bin/pioMESHTASTIC_FIRMWARE_ROOT)esptool, nrfutil, picotool on $PATH (or under the firmware venv at .venv/bin/) if you want to use the direct-tool wrapperscd <firmware-repo>/mcp-server
python3 -m venv .venv
.venv/bin/pip install -e .
Verify:
MESHTASTIC_FIRMWARE_ROOT=<firmware-repo> .venv/bin/python -m meshtastic_mcp
The server blocks on stdin (that's correct — it speaks MCP over stdio). Ctrl-C to exit.
Edit ~/.claude/settings.json (global) or <firmware-repo>/.claude/settings.local.json (project-only):
{
"mcpServers": {
"meshtastic": {
"command": "<firmware-repo>/mcp-server/.venv/bin/python",
"args": ["-m", "meshtastic_mcp"],
"env": {
"MESHTASTIC_FIRMWARE_ROOT": "<firmware-repo>"
}
}
}
}
Replace <firmware-repo> with the absolute path, e.g. /Users/you/GitHub/firmware. Restart Claude Code after editing.
Same mcpServers block, but in ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows).
| Tool | What it does |
|---|---|
list_devices | USB/serial port listing, flags likely-Meshtastic candidates |
list_boards | PlatformIO envs with custom_meshtastic_* metadata; filters by arch/supported/query/level |
get_board | Full env dict incl. raw pio config |
| Tool | What it does |
|---|---|
build | pio run -e <env> (+ mtjson target) |
clean | pio run -e <env> -t clean |
pio_flash | pio run -e <env> -t upload --upload-port <port> — any architecture |
erase_and_flash | ESP32 full factory flash via bin/device-install.sh |
update_flash | ESP32 OTA app-partition update via bin/device-update.sh |
touch_1200bps | 1200-baud open/close to trigger USB CDC bootloader entry |
Backed by long-running pio device monitor subprocesses with a 10k-line ring buffer per session and board-specific filters (esp32_exception_decoder auto-selected when you pass env=).
| Tool | What it does |
|---|---|
serial_open | Start a monitor session; returns session_id |
serial_read | Cursor-based pull; reports dropped if lines aged out of the ring |
serial_list | All active sessions |
serial_close | Terminate a session |
| Tool | What it does |
|---|---|
device_info | my_node_num, long/short name, firmware version, region, channel, node count |
list_nodes | Full node database with position, SNR, RSSI, last_heard, battery |
The tool tables below document 38 currently registered MCP server tools.
| Tool | What it does |
|---|---|
set_owner | Long name + optional short name (≤4 chars) |
get_config | One section or all (LocalConfig + ModuleConfig) |
set_config | Dot-path field write: lora.region="US", device.role="ROUTER", etc. |
get_channel_url | Primary-only or include_all=admin URL |
set_channel_url | Import channels from a Meshtastic URL |
set_debug_log_api | Enable or disable debug logging for the Meshtastic Python API client |
send_text | Broadcast or direct text message |
reboot | localNode.reboot(secs) — requires confirm=True |
shutdown | localNode.shutdown(secs) — requires confirm=True |
factory_reset | localNode.factoryReset(full?) — requires confirm=True |
| Tool | What it does |
|---|---|
esptool_chip_info | Read chip, MAC, crystal, flash size |
esptool_erase_flash | Full-chip erase (destructive) |
esptool_raw | Pass-through; confirm=True required for write/erase/merge |
nrfutil_dfu | DFU-flash a .zip package |
nrfutil_raw | Pass-through |
picotool_info | Read Pico BOOTSEL-mode info |
picotool_load | Load a UF2 |
picotool_raw | Pass-through |
| Tool | What it does |
|---|---|
uhubctl_list | Enumerate USB hubs + attached-device VID/PID (read-only) |
uhubctl_power | Drive a hub port on or off; off requires confirm=True |
uhubctl_cycle | Off → wait delay_s → on; confirm=True required |
Target a port by explicit (location, port) (raw uhubctl syntax like
location="1-1.3", port=2) or by role ("nrf52", "esp32s3"). Role
lookup checks MESHTASTIC_UHUBCTL_LOCATION_<ROLE> +
MESHTASTIC_UHUBCTL_PORT_<ROLE> env vars first, then auto-detects via VID
against uhubctl's output.
Requires uhubctl on PATH:
brew install uhubctl # macOS
apt install uhubctl # Debian/Ubuntu
Modern macOS + PPPS-capable hubs generally work without root. On Linux
without udev rules, or on old macOS with driver quirks, you may need
sudo. If uhubctl returns a permission error the MCP tool raises a
clear UhubctlError pointing at the
udev-rules / sudo fallback
rather than auto-sudo'ing mid-run.
confirm=True as a tool-level gate, on top of any permission prompt from Claude.serial_* session is active on a port, device_info/admin tools on the same port will fail fast with a pointer at the active session_id. Close the session first.erase_and_flash / update_flash error if the env's architecture isn't ESP32 — use pio_flash for nRF52/RP2040/STM32.| Var | Default | Purpose |
|---|---|---|
MESHTASTIC_FIRMWARE_ROOT | walks up from cwd for platformio.ini | Pin the firmware repo |
MESHTASTIC_PIO_BIN | ~/.platformio/penv/bin/pio → $PATH pio → platformio | Override pio location |
MESHTASTIC_ESPTOOL_BIN | <firmware>/.venv/bin/esptool → $PATH | Override esptool |
MESHTASTIC_NRFUTIL_BIN | $PATH | Override nrfutil |
MESHTASTIC_PICOTOOL_BIN | $PATH | Override picotool |
MESHTASTIC_MCP_SEED | mcp-<user>-<host> | PSK seed for test-harness session (CI override) |
MESHTASTIC_MCP_FLASH_LOG | <mcp-server>/tests/flash.log | Tee target for pio/esptool/nrfutil subprocess output (TUI tails it) |
MESHTASTIC_MCP_TCP_HOST | unset | host or host:port of a meshtasticd daemon to surface as a TCP device (see "TCP / native-host nodes" below) |
The native-macos and native PlatformIO envs build a headless meshtasticd
binary that runs on the host (Apple Silicon / Intel macOS, or Linux Portduino).
The daemon exposes the meshtastic TCP API on port 4403 rather than a USB
serial endpoint — point the MCP server at it via MESHTASTIC_MCP_TCP_HOST:
# 1. Build + run a daemon on this host (see variants/native/portduino/platformio.ini
# for full Homebrew prereqs and CH341 LoRa-adapter setup).
pio run -e native-macos
~/.meshtasticd/meshtasticd
# 2. Point the MCP server at it.
export MESHTASTIC_MCP_TCP_HOST=localhost # or host:port, default port 4403
First-run gotcha — MAC address. meshtasticd derives its MAC from the
USB adapter's serial-number / product strings. Many cheap CH341 dongles
(MeshStick included — VID 0x1A86 / PID 0x5512) ship with iSerialNumber=0
and iProduct=0, so the daemon aborts on boot with *** Blank MAC Address not allowed!. Set the MAC explicitly in config.yaml:
# Under General:
MACAddress: 02:CA:FE:BA:BE:01
Use a locally-administered address (first byte's second-LSB set, e.g.
02:* / 06:* / 0A:* / 0E:*) to avoid colliding with a real OUI.
There is also a --hwid AA:BB:CC:DD:EE:FF CLI flag visible in
meshtasticd --help, but it is currently broken in
MAC_from_string() (src/platform/portduino/PortduinoGlue.cpp): the
function strips colons from its parameter but then reads bytes from the
global portduino_config.mac_address, so --hwid is silently overridden
when MACAddress: is also set, and crashes the daemon (uncaught
std::invalid_argument: stoi: no conversion) when it isn't. Use the YAML
form until that's fixed upstream.
list_devices will surface the daemon as tcp://localhost:4403 with
likely_meshtastic=True, so device_info, list_nodes, get_config,
set_config, set_owner, send_text, userprefs_*, and the admin RPCs
auto-select it when no port is passed. Pass port="tcp://other-host:9999"
explicitly to target a different daemon.
Tools that don't apply to a TCP/native node (no USB hardware to operate
on) raise a clear ConnectionError rather than failing mysteriously:
pio_flash, erase_and_flash, update_flash, touch_1200bps,
serial_open (use info/admin tools directly), and the vendor escape hatches
esptool_*, nrfutil_*, picotool_*. pio_flash against a native* env
similarly raises — there's no upload step; use build and run the binary
directly.
The pytest harness in tests/ still assumes USB-attached devices per role —
TCP-aware fixtures are not part of this surface yet.
mcp-server/tests/ holds a pytest-based integration suite that exercises
real USB-connected Meshtastic devices against the MCP server surface. Separate
from the native C++ unit tests in the firmware repo's top-level test/
directory — this one validates the device-facing behavior end-to-end.
./mcp-server/run-tests.sh # full suite (auto-detect + auto-bake-if-needed)
./mcp-server/run-tests.sh --force-bake # reflash devices before testing
./mcp-server/run-tests.sh --assume-baked # skip the bake step (caller vouches for state)
./mcp-server/run-tests.sh tests/mesh # one tier
./mcp-server/run-tests.sh tests/mesh/test_traceroute.py # one file
./mcp-server/run-tests.sh -k telemetry # pytest name filter
The wrapper auto-detects connected devices (VID 0x239A → nrf52 → env
rak4631; 0x303A or 0x10C4 → esp32s3 → env heltec-v3), exports
MESHTASTIC_MCP_ENV_<ROLE> env vars, and invokes pytest. Overrides via
per-role env vars: MESHTASTIC_MCP_ENV_NRF52=heltec-mesh-node-t114 ./run-tests.sh.
No hardware connected? The wrapper narrows to tests/unit/ only and says so
in the pre-flight header.
bake (tests/test_00_bake.py) — flashes both hub roles with the
session's test profile. Has a skip-if-already-baked check (region + channel
match); --force-bake overrides.unit — pure Python, no hardware. boards / PIO wrapper /
userPrefs-parse / testing-profile fixtures.mesh — 2-device mesh: formation, broadcast delivery, direct+ACK,
traceroute, bidirectional. Parametrized over both directions. Includes
test_peer_offline_recovery which uses uhubctl to power-cycle one peer
mid-conversation and verifies the mesh recovers (skips without uhubctl).telemetry — periodic telemetry broadcast + on-demand request/reply
(TELEMETRY_APP with wantResponse=True).monitor — boot log has no panic markers within 60 s of reboot.recovery — uhubctl power-cycle round-trip: verifies the hub port
can be toggled off/on, the device re-enumerates with the same
my_node_num, and NVS-resident config (region, channel, modem preset)
survives a hard reset. Requires uhubctl on PATH; skips cleanly otherwise.ui — input-broker-driven screen navigation (AdminMessage.send_input_event
injection → Screen::handleInputEvent → frame transition). Parametrized
on the screen-bearing role (heltec-v3 OLED). Captures images via USB
webcam + OCRs them for HTML-report evidence. Requires pip install -e '.[ui]'
and MESHTASTIC_UI_CAMERA_DEVICE_ESP32S3=<index>; tier is auto-deselected
if cv2 isn't importable.fleet — PSK-seed isolation: two labs with different seeds never
overlap.admin — owner persistence across reboot, channel URL round-trip,
lora.hop_limit persistence.provisioning — region/channel baking, userPrefs survive
factory_reset(full=False).The tests/ui/ tier drives the on-device OLED via the firmware's existing
AdminMessage.send_input_event RPC (no firmware changes required) and
verifies transitions via a macro-gated log line + camera + OCR. Summary:
pip install -e 'mcp-server/.[ui]' — pulls in
opencv-python-headless, numpy, easyocr, Pillow. First easyocr
run downloads ~100 MB of models to ~/.EasyOCR/; an autouse session
fixture pre-warms the reader so per-test OCR is <100 ms after that..venv/bin/python -c "import cv2; [print(i, cv2.VideoCapture(i).read()[0]) for i in range(5)]"
export MESHTASTIC_UI_CAMERA_DEVICE_ESP32S3=0
./run-tests.sh tests/ui -v
tests/ui_captures/<session_seed>/<test_id>/, one
PNG + .ocr.txt per frame_capture() call, with a per-test
transcript.md stepping through event → frame → OCR. The HTML report
embeds the full image strip inline (pass or fail).On macOS, cv2.VideoCapture(0) triggers the TCC Camera permission prompt
on first use. Pre-grant Terminal (or your IDE's terminal) before running.
The OpenCVBackend fails fast on 10 consecutive black frames so a silent
permission denial surfaces as a clear error, not an empty PNG strip.
No camera? Set MESHTASTIC_UI_CAMERA_BACKEND=null (or leave the device var
unset). Tests still exercise the event-injection path and log assertions;
captures just become 1×1 black PNGs.
tests/)report.html — self-contained pytest-html report. Each test gets a
Meshtastic debug section attached on failure with a 200-line firmware
log tail + device-state dump. Open this first on failures.junit.xml — CI-parseable.reportlog.jsonl — pytest-reportlog event stream; consumed by the TUI.fwlog.jsonl — firmware log mirror (meshtastic.log.line pubsub → JSONL).flash.log — tee of all pio / esptool / nrfutil / picotool subprocess
output during the run (driven by MESHTASTIC_MCP_FLASH_LOG)..venv/bin/meshtastic-mcp-test-tui
.venv/bin/meshtastic-mcp-test-tui tests/mesh # pytest args pass through
Textual-based wrapper over run-tests.sh with a live test tree, tier
counters, pytest output pane, firmware-log pane, and a device-status strip.
Key bindings: r re-run focused, f filter, d failure detail, g open
report.html, x export reproducer bundle, l cycle fw-log filter, q
quit (SIGINT → SIGTERM → SIGKILL escalation).
Set MESHTASTIC_UI_TUI_CAMERA=1 to mount a bottom-of-screen UI camera
panel. Left side: the latest capture PNG rendered as Unicode half-blocks
(via rich-pixels, works in any terminal — no kitty/sixel required).
Right side: live transcript tail ("step 3 — frame 4/8 name=nodelist_nodes
— OCR: Nodes 2/2") so you can see every event-injection and its result
as each UI test runs. Requires the [ui] extras for image rendering; the
transcript alone works without them.
Three AI-assisted workflows are wired up for Claude Code operators
(.claude/commands/) and Copilot operators (.github/prompts/):
/test (run + interpret), /diagnose (read-only health report), /repro
(flake triage, N-times re-run with log diff).
tests/conftest.py snapshot + restore
userPrefs.jsonc; never edit userPrefs.jsonc from inside a test.
Use the test_profile / no_region_profile fixtures for ephemeral
overrides.SerialInterface holds an exclusive port lock; sequence calls
open → mutate → close, then next device. No parallel calls to the
same port.tests/mesh/_receive.py::nudge_nodeinfo_port and the three directed-
send tests (test_direct_with_ack, test_traceroute,
test_telemetry_request_reply) for the canonical pattern.mcp-server/
├── pyproject.toml
├── README.md
└── src/meshtastic_mcp/
├── __main__.py # entry: python -m meshtastic_mcp
├── server.py # FastMCP app + @app.tool() registrations (thin)
├── config.py # firmware_root, pio_bin, esptool_bin, etc.
├── pio.py # subprocess wrapper (timeouts, JSON, tail_lines)
├── devices.py # list_devices (findPorts + comports)
├── boards.py # list_boards / get_board (pio project config parse + cache)
├── flash.py # build, clean, flash, erase_and_flash, update_flash, touch_1200bps
├── serial_session.py # SerialSession + reader thread + ring buffer
├── registry.py # session registry + per-port locks
├── connection.py # connect(port) ctx mgr — SerialInterface + port lock
├── info.py # device_info, list_nodes
├── admin.py # set_owner, get/set_config, channels, send_text, reboot/shutdown/factory_reset
└── hw_tools.py # esptool / nrfutil / picotool wrappers
MESHTASTIC_FIRMWARE_ROOT.pio" — install PlatformIO or set MESHTASTIC_PIO_BIN.serial_close(session_id) or serial_list to find it.factory.bin not found after build — the env may not be ESP32; only ESP32 envs produce a .factory.bin.touch_1200bps reported new_port: null — the device may not have 1200bps-reset stdio, or the bootloader re-uses the same port name. Check list_devices manually.