Jean-Pierre MartineauandClaude Opus 4.8
421781f13d
Trace: - Auto-path now ends every route at the target's own hash (verified on hardware: a trace only replies when its destination is the final hop), so direct neighbors and learned multi-hop routes both resolve correctly. - Retry the transient "no event received during contacts retrieval" blip and surface a clean DeviceCommandError instead of a raw traceback. - Route summary annotates each node with its hash at the command's path-hash width, wrapping after an arrow when the line is too long. - Default trace count changed from 5 to 3. Also includes the in-progress device-config feature (config_io, device_config, config tool, config editor UI, tests) and related supporting edits. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
MeshTools
A modern, extensible command-line toolkit for tuning and exploring a MeshCore mesh through a serial-connected companion device.
MeshTools is interactive by default (a Rich + Questionary menu) and fully scriptable (every menu option is also a Typer subcommand). Every run is logged to SQLite and can be turned into interactive HTML visualizations.
Features
| Feature | Status | What it does |
|---|---|---|
| Trace | ✅ skeleton | Run repeated path traces to a target and aggregate per-hop SNR. |
| Device info | ✅ skeleton | Show the connected companion device's identity and radio config. |
| Device config | ✅ working | View and change every setting (name, radio, behavior, experimental), with TOML backup/restore, channels, and gated destructive ops. |
| Device discovery | ✅ working | Enumerate serial devices, pick one interactively, and remember the last good default. |
| History / DB | ✅ skeleton | Every tool execution and measurement is persisted and queryable. |
| TX-power optimization | ✅ working | Sweep transmit power (coarse + refine), converge on the best signal, optionally apply, render an interactive chart. |
| Path optimization | 🚧 planned | Discover and score mesh paths from neighbors outward. |
| Visualizations | 🟡 partial | Plotly TX-sweep chart (HTML) done; pyvis mesh graph + time-series planned. |
Install
pip install -e ".[dev]"
Python 3.10+ is required (matching the meshcore library).
Usage
# Interactive menu (default)
meshtools
# Scripted: every menu option is also a subcommand
meshtools trace --target Alice --samples 10 --profile yagi
# Force a route through specific repeaters (MeshCore-app style): comma-separated
# contact names and/or hex key prefixes, mixed freely. Blank lets the device route.
meshtools trace --target Alice --path "3d,f2,3d"
meshtools trace --target Alice --path "3d,Bravo-Repeater,f2"
meshtools tx-optimize --target Alice --samples 6 --step 3 --apply
meshtools info
meshtools history
# Device configuration: view, set, back up, restore
meshtools config # show all current settings
meshtools config set radio_sf 9 # change one setting
meshtools config backup node.toml # archive every setting to TOML
meshtools config restore node.toml --dry-run
# No radio attached? Use the built-in simulator for development.
meshtools --mock trace --target Alice
Architecture
MeshTools is layered so features are added by dropping a single file in meshtools/tools/:
cli.py Typer app; no subcommand -> interactive menu
context.py AppContext (console, config, repository, device) dependency container
core/ Domain: models, config + device profiles, connection abstraction (+ mock)
tools/ Pluggable "menu options"; each self-registers and gets logging for free
services/ Algorithms (trace aggregation, tx search, path exploration) — no UI
persistence/ SQLite schema, repository, structured logging
ui/ Rich theme/widgets + Questionary menu
viz/ Plotly / pyvis renderers -> HTML & PNG
Adding a feature means subclassing Tool, decorating it with @register, and
implementing run(). The same registry builds both the CLI and the menu, and the base
class wraps every execution in a logged runs row automatically.