AI tools are changing how we use software — but they still cover only a fraction of our work. Our daily tasks are scattered across browsers, desktop apps, terminals, and messaging tools — each with its own interface and habits, disconnected from each other.

Understudy is a teachable desktop agent. It operates your computer like a human colleague — GUI, browser, shell, file system, all in one local runtime. You show it a task once, it extracts the intent (not just the coordinates), remembers the successful path, discovers faster execution routes over time, and eventually handles routine work on its own. No API integrations required. No workflow builders. Just demonstrate once.

Understudy is designed as a layered progression — the same journey a new hire takes when they grow into a reliable colleague.

Day 1:    Watches how things are done
Week 1:   Imitates the process, asks questions
Month 1:  Remembers the routine, does it independently
Month 3:  Finds shortcuts and better ways
Month 6:  Anticipates needs, acts proactively

That's why it's called Understudy — in theater, an understudy watches the lead, learns the role, and steps in when needed.

Each of the five layers maps to a stage of this journey:

Layer 1 ┃ Operate Software Natively     Operate any app a human can — see, click, type, verify
────────╋──────────────────────────────────────────────────────────────────────────────────
Layer 2 ┃ Learn from Demonstrations     User shows a task once — agent extracts intent, validates, learns
────────╋──────────────────────────────────────────────────────────────────────────────────
Layer 3 ┃ Crystallized Memory           Agent accumulates experience from daily use, hardens successful paths
────────╋──────────────────────────────────────────────────────────────────────────────────
Layer 4 ┃ Route Optimization            Automatically discover and upgrade to faster execution routes
────────╋──────────────────────────────────────────────────────────────────────────────────
Layer 5 ┃ Proactive Autonomy            Notice and act in its own workspace, without disrupting the user

Current status: Layers 1-2 are implemented and usable today. Layers 3-4 are partially implemented. Layer 5 is still the long-term direction.

Every layer depends on the one below it. No shortcuts — the system earns its way up. Read the full story: Overview → | Chinese Overview → | Product Design →

Demo environment: macOS + GPT-5.4 via Codex. See Supported Models for all providers.

The video has been sped up. Full-length original: Google Drive. For inline playback, visit the Overview page.

Status: implemented today on macOS.

Understudy is not just a GUI clicker. It's a unified desktop runtime that mixes every execution route your computer offers — in one agent loop, one session, one policy pipeline:

The planner decides which route to use for each step. A single task might browse a website, run a shell command, click through a native app, and send a message — all in one session.

GUI grounding — dual-model architecture: the main model decides what to do, a separate grounding model decides where on screen. HiDPI-aware across Retina displays, with automatic high-res refinement for small targets and two grounding modes (simple prediction or multi-round validation with simulation overlays). Grounding benchmark: 30/30 targets resolved — explicit labels, ambiguous targets, icon-only elements, fuzzy prompts.

For implementation details (coordinate spaces, stabilization, cross-attempt feedback, capture modes), see Product Design.

Status: implemented today. Teach can record a demo, analyze it, clarify the task, optionally validate the replay, and publish a reusable skill.

Explicit teaching: you deliberately show the agent how to do a task. Not a macro recorder — Understudy learns intent, not coordinates.

/teach start                              Start dual-track recording (screen video + semantic events)
                                          → Perform the task once with mouse and keyboard

/teach stop "file the weekly expense report"   AI analyzes the recording → extracts intent, parameters,
                                               steps, success criteria → creates a teach draft
                                               → Multi-turn dialogue to refine the task

/teach confirm [--validate]               Lock the task card; optionally replay to validate

/teach publish <draftId> [skill-name]     Generate SKILL.md → hot-load into active sessions

The published SKILL.md is a three-layer abstraction: intent procedure (natural language steps), route options (preferred / fallback paths), and GUI replay hints (last resort only, re-grounded from the current screenshot each time). UI redesigns, window resizing, even switching to a similar app — as long as the semantic target still exists, the skill works.

A real published output example is available at examples/published-skills/taught-person-photo-cutout-bc88ec/SKILL.md. It is kept under examples/ on purpose, so the repo documents the artifact format without auto-loading it as a real workspace skill.

For the full teach pipeline, evidence pack construction, and validation details, see Product Design.

Status: partially implemented. Understudy now has a working workflow crystallization loop for repeated day-to-day use, but promotion policy and automatic route upgrading are still early.

Implicit learning: as you use Understudy day-to-day, it automatically identifies recurring patterns and crystallizes successful paths — no explicit teaching needed:

What the user experiences today

• You use Understudy normally; no /teach command is required.
• Repeated multi-turn work in the same workspace is compacted in the background.
• Once a pattern repeats often enough, Understudy publishes a workspace skill automatically, hot-loads it into active sessions, and sends a notification that a crystallized workflow is ready.
• From that point on, the agent can pick the new skill through the normal ## Skills (mandatory) flow, just like a taught skill.

Current boundaries

• The learning loop is intentionally conservative: if the pattern is not clear enough, Understudy keeps exploring instead of pretending it has learned it.
• Segmentation, clustering, and skill synthesis are currently LLM-first.
• Promotion thresholds are still heuristic rather than fully policy-driven.
• Layer 4 style route upgrading is still limited; crystallization mainly produces reusable skills today rather than fully automatic route replacement.

For the implementation details (crystallization pipeline, segmentation, clustering, skill synthesis), see Product Design.

Status: partially implemented. Route preferences, route guard policies, and teach route annotations are working today. Fully automatic route discovery and promotion are still in progress.

The same feature can be accomplished in multiple ways. Take "send a Slack message" as an example:

Fastest ──→  1. API call         Hit the Slack API directly (milliseconds)
             2. CLI tool         slack-cli send (seconds)
             3. Browser          Type and send in Slack's web app (seconds)
Slowest ──→  4. GUI              Screenshot-ground, click, type in the Slack desktop client (seconds~10s)

Starts at GUI on Day 1 — because GUI is the universal fallback that works with any app. As usage accumulates, Understudy progressively discovers faster ways to accomplish the same feature, upgrading after safe verification.

What the user experiences today

• The system prompt explicitly steers the agent toward faster routes: direct tool/API > Shell/CLI > browser > GUI.
• A route guard tracks consecutive failures per route. If one route keeps failing (e.g., GUI grounding is unreliable for a particular app), the agent is prompted to switch to an alternative.
• Teach-produced skills annotate each step with preferred / fallback / observed routes, so the agent can skip GUI when a faster path is available.
• Browser mode auto-falls back: tries Chrome extension relay first, switches to managed Playwright if that fails.
• The GUI capability matrix dynamically enables/disables tool subsets based on available permissions, so the agent never attempts a route it can't execute.

What this looks like in practice

• A newly learned task may start with GUI as the observed fallback path.
• Over time, the agent starts preferring higher-level routes — an existing skill, browser flow, or shell command — when those preserve the same result.
• In practice this usually feels like fewer brittle GUI steps, faster execution, and fewer repeated detours.
• When a better route is uncertain, the agent falls back to the slower but safer path rather than risk a failure.

Current boundaries

• Route optimization today is guidance and safe preference ordering, not a fully autonomous optimizer.
• The agent can prefer a known faster route, but does not yet run a broad automatic search over all possible routes for every recurring task.
• Automatic route promotion (discover → verify → promote to default) is designed but not yet fully implemented.

For the full route selection mechanisms, upgrade policy, and future direction, see Product Design.

Status: mostly vision today. Scheduling and runtime surfaces exist, but passive observation, isolated workspace, and proactive autonomy are still ahead.

Earned, not assumed. Understudy's ultimate goal isn't "wait for instructions" — it's becoming a colleague that observes long-term, acts proactively.

Long-term observation and learning — beyond explicit teach, Understudy passively observes your daily operations, analyzes recurring patterns, and understands your work habits and preferences. It's not recording every click — it's understanding what you do each day, when you do it, and which tools you use.

Proactive suggestions — based on accumulated observations, proactively suggests what to do next: emails to process, tasks to follow up on, reports to run. Suggestions are presented non-intrusively; only executed after you confirm.

Isolated workspace — executes tasks in its own desktop, without occupying your screen:

Cross-app orchestration — the isolated workspace unlocks true multi-app parallel operation. In its own desktop, Understudy can control multiple apps simultaneously, coordinating data flow between them — extract info from email into a spreadsheet, update the calendar, and send results via Slack. No longer a single-threaded, one-window-at-a-time model.

Progressive trust model — each skill starts at manual, promoted only through sustained success:

 ┌──────────┐ ┌───────────┐ ┌─────────┐ ┌──────────┐ ┌─────────┐
 │ Terminal  │ │ Dashboard │ │ WebChat │ │ Telegram │ │  Slack  │ ...
 └────┬─────┘ └─────┬─────┘ └────┬────┘ └────┬─────┘ └────┬────┘
      └──────────────┴────────────┴────────────┴────────────┘
                                  │
                    Gateway (HTTP + WebSocket + JSON-RPC)
                                  │
                     ┌────────────┴────────────┐
                     │                         │
              Session Runtime            Built-in Tools
              + Policy Pipeline    ┌──────────────────────┐
                                   │ gui_*  │ browser     │
                                   │ bash   │ memory      │
                                   │ web    │ schedule    │
                                   │ message│ subagents   │
                                   └──────────────────────┘

• Local-first — screenshots, recordings, and traces are stored locally by default; GUI grounding and teach analysis may send selected images or keyframes to your configured model provider
• One gateway — terminal, web, mobile, messaging apps all connect through one endpoint
• One session runtime — chat, teach, scheduled tasks, subagents share the same loop
• Policy pipeline — every tool call passes through safety, trust, and logging hooks

npm install -g @understudy-ai/understudy
understudy wizard    # walks you through setup

git clone https://github.com/understudy-ai/understudy.git
cd understudy
pnpm install && pnpm build
pnpm start -- wizard

# Recommended: start the gateway daemon, then use the terminal
understudy daemon --start        # Start the gateway background process (or: understudy gateway --port 23333)
understudy chat                  # Terminal interactive mode (auto-connects to running gateway)

# Management UI
understudy dashboard             # Open the control panel in your browser

# Other entry points
understudy webchat               # Browser chat interface
understudy agent --message "..."  # Script/CI single-turn call (requires gateway)

Required for all installs:

Required for source installs / development:

macOS GUI automation (required for GUI tools and teach-by-demonstration on macOS):

Optional:

Understudy is currently developed and tested on macOS. Core features (CLI, gateway, browser, channels) are cross-platform by design, but native GUI automation and teach-by-demonstration require macOS today. Linux and Windows GUI backends are planned — contributions welcome.

Understudy is model-agnostic. Configure models with provider/model (for example anthropic/claude-sonnet-4-6).

The source of truth is the model catalog bundled with your current Understudy runtime plus any custom entries in your local model registry. That catalog evolves over time, so this README intentionally avoids freezing an exhaustive provider/model matrix.

Use one of these to see what your install can actually use:

understudy models --list
understudy wizard

Auth and provider notes:

Default: openai-codex/gpt-5.4. The wizard chooses from models detected in your local runtime auth/model registry.

Full screenshot-grounded GUI automation on macOS depends on the native helper plus two system permissions. Missing permissions degrade or hide parts of the GUI toolset rather than failing every GUI capability equally.

What it enables: Mouse clicks, typing, drag, scroll, hotkeys against app targets, and demonstration event capture.

If missing: Input-driving tools are blocked. Observation-oriented tools such as screenshots and GUI reads may still work when Screen Recording is available.

How to grant:

1. Open System Settings → Privacy & Security → Accessibility
2. Click the + button, add your terminal app (Terminal.app, iTerm2, VS Code, etc.)
3. Toggle it on

What it enables: Screenshots for GUI grounding and verification, screenshot-based reads, and demonstration video recording.

If missing: Screenshot-grounded tools are blocked. Keyboard-only tools such as gui_keypress and gui_hotkey can still remain available, and some tools can still operate in targetless mode.

How to grant:

1. Open System Settings → Privacy & Security → Screen Recording
2. Click the + button, add your terminal app
3. Toggle it on
4. Restart your terminal app for the permission to take effect

Both permissions must be granted to the terminal application that runs understudy, not to Understudy itself.

If GUI behavior still looks partial after granting permissions, run understudy doctor --deep to check the native helper, grounding availability, OCR, browser runtime, and teach-analysis dependencies.

47 built-in skills + workspace skills you create or install.

understudy skills --list                    # browse available skills
understudy skills install <name-or-url>     # install from registry or URL

# Or teach by demonstration
/teach start → demonstrate → /teach stop → /teach confirm
→ /teach publish <draftId>
# Optional: /teach validate <draftId> before or after publish

8 built-in channel adapters: Web, Telegram, Discord, Slack, WhatsApp, Signal, LINE, iMessage.

understudy channels --list
understudy channels --add telegram

apps/cli           CLI entrypoints, 30+ operator commands
packages/core      Agent session runtime, config, auth, skills, policies
packages/gateway   HTTP + WebSocket gateway, session runtime, web surfaces
packages/gui       Native GUI runtime, screenshot grounding, demo recorder
packages/tools     Built-in tools: browser, web, memory, schedule, GUI, message
packages/channels  Channel adapters (8 platforms)
packages/types     Shared TypeScript type definitions
skills/            47 built-in skill modules
examples/          Sample published artifacts and reference outputs
docs/              Vision, product design documentation

pnpm install          # install dependencies
pnpm build            # build all packages
pnpm test             # run tests
pnpm lint             # lint with oxlint
pnpm typecheck        # type-check all packages
pnpm check            # full validation: build + lint + typecheck + test

Understudy builds on ideas and code from several outstanding open-source projects:

Special thanks to Mario Zechner for the pi-agent-core foundation that powers Understudy's agent loop.

See CONTRIBUTING.md for guidelines.

We're looking for contributors in these areas:

• GUI backends — Linux (AT-SPI) and Windows (UIA) native GUI support
• Skills — new skill modules for popular apps and workflows
• Route discovery — automatic API detection and upgrade logic (Layer 4)
• Teach improvements — better evidence pack analysis and validation
• Documentation & translations

MIT