Scoped desktop capture
Capture one selected desktop application in an attended session, including visible same-process top-level windows and owned menus, dialogs, and tool popups whose root-owner chain leads to the selected application.
UI Atlas
UI Atlas is an offline-first Windows toolkit for manually recording a selected desktop application and its owned popups, then building inspectable UI knowledge graph layers. It helps developers, researchers, and QA teams preserve structured interface evidence and export maps without uploading recordings or requiring an account.
Offline-first mapping for Windows application interfaces
UI Atlas is an offline-first Windows toolkit for manually recording a selected desktop application and its owned popups. It preserves raw capture evidence, then deterministically builds Raw World and Semantic World UI Knowledge Graph layers that can be inspected, compared, validated, and exported.
The project is intended to make unfamiliar Windows interfaces easier to analyze by documenting screens, controls, interactions, and resulting states without taking actions on the user’s behalf. Its v1 workflow is deliberately bounded: a user selects a target, records attended clicks and observations, builds a map, reviews quality and graph layers, and consumes the resulting data through an independent library or the command-line tooling.
Features
Immutable recording bundles
Preserve ordered input events, interaction attempts, outcomes, diagnostics, capture health, observations, and optional scoped application or owned-popup frames as raw evidence.
Two-layer UI knowledge graph
Build Raw World and Semantic World UI Knowledge Graph layers from recordings, retaining lineage links between windows, variants, controls, and preceding worlds.
Multi-lane observation
Use independent pixels, UI Automation, visual geometry, OCR, and bounded native-verification lanes where the target permits them; screenshot failures can degrade capture without discarding other evidence.
Inspection and quality review
Validate recordings and maps, generate quality reports, inspect world streams and graph layers, and review maps in a read-only WPF explorer.
Structured exports
Export identity-bearing map data as human-readable JSON, or use the documented ui-atlas_flat and SQLite map formats for downstream consumption.
Use cases
Reverse-engineer an unfamiliar desktop UI
A developer can record baseline and post-click observations of an unfamiliar Windows application, then inspect the resulting windows, controls, routes, and state transitions in a structured map.
Create reproducible UI evidence
A researcher or QA engineer can preserve screenshots, Win32 metadata, bounded UI Automation observations, input timing, and interaction outcomes as a replayable evidence bundle for review.
Feed structured interface data to another system
A team building an AI interface-understanding system can consume the lossless JSON map export, including interaction sessions, merged routes, affordances, negative examples, and lineage data.
Review and qualify captured maps
A maintainer can run map validation and quality checks, inspect Raw World and Semantic World layers, and use the read-only explorer to identify duplicate screens, partial scans, or clicks without a confirmed result.
Pros and Cons
Pros
- Keeps recordings offline and does not require an account or upload of recordings.
- Preserves immutable raw evidence and supports recoverable partial or cancelled bundles when retention policy permits.
- Combines screenshots and structured Windows/UI Automation evidence, with validation, quality review, inspection, and multiple export options.
- Provides separate Raw World and Semantic World layers with lineage back to observed evidence.
Cons
- The supported v1 scope is limited to Windows 10 version 2004 or later and Windows 11, with a required .NET runtime for release installations.
- Capture is attended and manually driven; the project does not provide application action execution, authoring, remote control, or background monitoring.
- The documented workflow targets one selected desktop application and its owned popups rather than a general multi-application monitoring environment.
FAQ
What platforms and prerequisites does UI Atlas support?
UI Atlas supports Windows 10 version 2004 or later and Windows 11. The release ZIP requires the .NET 10 Windows Desktop Runtime; source builds require the .NET SDK 10.0.200.
How is UI Atlas installed?
Extract the release ZIP and run install.ps1 from the extracted ui-atlas-core folder. Installation is per-user, does not require administrator rights, and creates a UiAtlas Recorder Start-menu shortcut. The extracted folder can also be run portably with ui-atlas.cmd.
What does a recording session capture?
The recorder captures one selected desktop application and its owned popups during an attended manual session. It takes a baseline observation, records each attended click and its resulting observation, and supports explicit finish or cancellation.
What outputs does UI Atlas produce?
Recordings use immutable .mlrec ZIP bundles. Generated maps can be stored as SQLite and exported as human-readable JSON, with additional map export formats including ui-atlas_flat and SQLite.
Does UI Atlas control applications or monitor them in the background?
No. The documented v1 scope does not include application actions, authoring, execution, remote control, or background monitoring. It is designed for attended manual capture and inspection.
Quick Facts
- Category
- Developer tool
- Platform
- Windows 10 version 2004 or later and Windows 11
- Primary workflow
- Attended manual recording of one selected desktop application
- Storage
- Local immutable .mlrec recordings, SQLite maps, JSON exports, and a recoverable catalog trash
- Interfaces
- Command-line tool and read-only WPF explorer
- Pricing
- No UiAtlas-specific pricing or licensing terms are documented in the supplied sources
Alternatives à UI Atlas
ByteAsk
ByteAsk is a terminal-first AI coding agent for C and C++ that edits repositories and verifies changes with the real compiler, debugger, sanitizers, and tests before showing a diff. It offers a free tier plus paid plans, with editor connectors and zero-retention handling described in the source.
Manta AI
Manta AI is an autonomous web app testing tool for teams that want to map application behavior, catch regressions, and generate tests without writing scripts or maintaining selectors. It works from a URL and supports plain-English test flows, run results with screenshots, and scheduled or deployment-triggered checks.
CreateOS Sandbox
CreateOS Sandbox is an isolated compute environment for running code and agent workloads inside Firecracker micro-VMs. It is designed for workflows that need machine-level isolation, private networking between sandboxes, and programmatic control through SDK, CLI, or MCP.
hob
hob is an independent workspace for coding agents that keeps agent sessions, terminals, history, and follow-up work organized around the tools and providers you already use. It is aimed at developers who want local control over routing, history, and workspace structure rather than a bundled model stack.
Ably Chat
Ably Chat is a chat API platform for building custom realtime chat applications. It supports room-based messaging, typing indicators, presence, reactions, and message updates, with usage-based pricing options for different deployment stages.
garden-md
garden-md is an open-source Node.js CLI that turns meeting transcripts into a local company wiki. It links entities across transcripts, keeps the original text intact, and renders the result as a browsable wiki.