Oliver Kieselbach dd7b10ace9 feat(agent): bandwidth estimate survives reboots via state persistence
The estimator accumulator previously lived in process memory only - a
mid-DeviceSetup reboot (e.g. a Win32 app forcing a restart) silently
discarded every rate sample collected so far. One enrollment commonly
spans several reboots.

- BandwidthStatePersistence (bandwidth-state.json in the state dir):
  mirrors the OfficeInstallStatePersistence pattern - atomic write via
  temp+rename, fail-soft load (missing/corrupt/foreign-session -> fresh
  start, never throws). SessionId-validated on load.
- BandwidthEstimator.ExportState/ImportState: sanitizes NaN/Infinity/
  negative samples, caps list sizes; baselines deliberately not
  persisted (a reboot gap always exceeds the per-file delta window).
- Save points are event-driven, never per-poll: the dormant transition
  (downloads just finished - exactly when app-forced reboots happen),
  after the interim emission, and at collector stop.
- The InterimEmitted guard rides in the same file: device_setup_end
  now fires once per SESSION, not once per process. Persisted AFTER
  the emission (StartupEventGate MarkEmitted ordering) so a crash
  between emit and save re-emits instead of suppressing forever.
- Wiring: DefaultComponentFactory passes the state directory through
  DeliveryOptimizationHost; the collector param is optional (null =
  in-memory only).

+13 tests (estimator roundtrip/merge/sanitizing, persistence roundtrip/
session-mismatch/corruption, collector-level resume incl. guard
survival); agent suite 1668/1668 green.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-09 11:55:45 +02:00
2026-07-06 21:50:51 +02:00
2026-03-26 07:02:33 +01:00
2026-07-06 20:53:00 +02:00

AutopilotMonitor

Status License Website

Advanced monitoring and troubleshooting solution for Windows Autopilot deployments. Gain full visibility into every enrollment session with a detailed event timeline, fleet health dashboards, and session reporting. Define custom analysis rules to automatically detect issues and gather rules to collect targeted evidence. Retrieve diagnostics packages on demand, configure agent settings like auto-reboot behavior and automatic timezone adjustment — all managed centrally from the web dashboard.

Private Preview

Autopilot Monitor is currently running as a Private Preview. Visit autopilotmonitor.com to request access and learn more.

Overview

Autopilot Monitor provides real-time tracking, intelligent diagnostics, and automated troubleshooting for Windows Autopilot enrollment processes. It consists of:

  • Bootstrap Script — PowerShell script deployed via Intune that starts monitoring early in the enrollment process
  • Monitoring Agent — Lightweight .NET application that collects telemetry and evidence during enrollment
  • Backend API — Azure Functions-based ingestion and processing pipeline
  • Web Dashboard — Next.js application for real-time monitoring and fleet analytics

Architecture

For detailed information about the system architecture, components, and data flow, see Architecture Documentation.

Documentation

Full documentation is available at autopilotmonitor.com/docs

License

This project uses a split licensing model:

  • MIT License — Agent (src/Agent/) and Shared library (src/Shared/) — unrestricted use on end-user devices
  • AGPL-3.0 — Backend (src/Backend/), Web Dashboard (src/Web/), and MCP Server (src/McpServer/) — server-side components remain open source

See LICENSE for full details.

S
Description
Windows Autopilot & Intune enrollment monitoring: live session timelines, fleet health, diagnostics, and an MCP server for AI-assisted troubleshooting.
Readme AGPL-3.0 60 MiB
Languages
C# 72.7%
TypeScript 25.7%
PowerShell 1%
JavaScript 0.5%