hwatu

Your agents are blind without hwatu

hwatu is a visual verification harness for coding agents, built as a WebKit daemon. Instead of "looks right to me", your agent gets one-call verified page checks in ~35 ms, pixel-diff scores it can climb, animations as numbers, and headless windows that never steal your focus, at any parallelism.

For human-in-the-loop tasks (e.g. Captcha), hwatu features a lightweight visual verification frontend renderer written in WebKit and a caller function. For tiling WMs (Hyprland, sway, niri, i3), hwatu is intended to replace your primary daily browser. Our current goal is to provide scrolling short-form content experience in mobile-level framerate.

Documents

• Vision: durable product principles, native platform strategy, swarm model
• Agent guide: protocol, primitives, verification loops
• Human guide: daily driving hwatu in a tiling WM, keybinds, media, hand-off
• Benchmarks: every number, measured, with methodology
• Roadmap: portfolio priorities and product boundaries
  • AI verification
  • Tiling-WM browser
  • Shared platform
• Continuous improvement: activation metric, feedback loop, weekly cadence
• Launch kit: reusable copy, channels, and measurement plan

Install

curl -fsSL https://raw.githubusercontent.com/hongnoul/hwatu/main/scripts/install.sh | bash

One static binary plus your distro's webkitgtk-6.0 (the installer checks). On Arch: yay -S hwatu. From source: cargo build --release.

Then pick your door, or take both:

hwatu setup             # agent: detect Claude Code, Cursor, Jcode, or MCP
hwatu localhost:3000    # human: open a window like you open a terminal

Real eyes for your coding agent

• STOP your agent claiming "pixel-perfect." Make it prove 97.49%.
• STOP paying 5 tool calls per page check. hwatu check is one call, ~35 ms (beats warm-server Playwright ~9x).
• STOP browser windows stealing your focus. Headless by default, you keep typing.
• STOP shipping 170 MB of Chromium. One static binary + your distro's webkitgtk.

hwatu setup detects supported coding agents and prints the available connections without changing their config. Choose a client explicitly when you are ready:

hwatu doctor
hwatu setup --client claude --scope project --dry-run
hwatu setup --client claude --scope project
hwatu demo

Setup is previewable, idempotent, and reversible with the same client and scope plus --undo. Manual MCP configuration remains one portable entry:

{ "mcpServers": { "hwatu": { "command": "hwatu", "args": ["mcp"] } } }

Or skip MCP entirely: every command is a short CLI call or one newline-delimited JSON line over a Unix socket.

Connecting hwatu makes its tools available; a project instruction tells the agent when to use them. Add this to AGENTS.md, CLAUDE.md, Cursor rules, or the equivalent for your harness:

## Frontend verification

Use Hwatu after frontend changes. Exercise the affected user journey and
verify its intended visible, navigational, or persisted result with `expect`.
A successful click or clean console is not proof of success. Check `console`
for additional JavaScript and request failures after verifying the outcome.

Then make the task's proof concrete:

Implement display-name editing on /settings. Use Hwatu to enter “Test User,”
save it, verify the visible success state, reload, confirm persistence, and
report any console errors.

The verification loop, real commands, real output:

hwatu --headless localhost:3000        # its window; you never see it
hwatu --headless staging.example.com   # the reference

hwatu diff --id 2 --other 1 --heatmap /tmp/heat.png
# {"match_percent":85.13,"regions":[{"x":0,"y":160,"w":2048,...}]}

hwatu motion --id 1                    # the reference's animations, as numbers
# easing cubic-bezier(0.25,1,0.5,1), 300ms, marquee 29.78px/s ...

# ...agent edits code...

hwatu diff --id 2 --other 1
# {"match_percent":97.49}              # climbing beats guessing

We ran this loop against a clone of stripe.com's landing page: an agent took it from 85.1% to 98.8% pixel match. Reproduce it: scripts/demo/. A second, real-agent scenario against AIUC (four responsive viewport diffs followed by live human hand-off) is reproducible with evidence manifests from scripts/demo-aiuc/.

A full verification pass (open, load, eval, screenshot, close) is one command, one tool call, ~35 ms median (benchmarks):

hwatu check localhost:5173 --eval 'document.title' --shot=/tmp/after.png
# {"title":"My App","eval":"My App","shot":"/tmp/after.png",
#  "console":[...],"load_ms":13,"total_ms":35}

Generated HTML in hand and no server? hwatu render is the same one-call pass with the markup as input: no temp file, no python3 -m http.server:

echo '<h1>generated</h1>' | hwatu render --stdin --shot=/tmp/gen.png
# {"rendered":true,"shot":"/tmp/gen.png","load_ms":5,"total_ms":28}

# React to load, console, download, and window events without polling.
hwatu watch --kinds load,console
# {"event":"load","seq":1,"window_id":7,"data":{"state":"started",...}}

MCP clients can call subscribe_events for the same stream as notifications/hwatu/event. See the full agent guide, including a larger copy-paste policy and verification loops.

Everywhere else, headless is decided at launch and a human can never see the session at any price. In hwatu it's a window property, switchable live, in both directions. And because hwatu is also the browser you already live in, the hand-off lands in a window that behaves like every other window on your desk, not a viewer bolted on for emergencies.

challenge is detection and hand-off only, by design: no solver APIs, no token injection, no fingerprint games.

Agents loop, you watch some reels

The hand-off works because hwatu is also a real browser, one built for tiling WMs. hwatu <url> opens a window like your terminal opens a shell (your WM is the tab bar, there is none in the window), with mainstream keybinds (ctrl+l, ctrl+f, ctrl+k palette, all rebindable via dotfile), native ad blocking (~119k EasyList rules compiled into WebKit's content-extension engine, zero JS in the request path), Chromium-curve scrolling, unmuted autoplay, a blur-shield that took Shorts from ~34 to ~95 fps, and one shortform control scheme (arrows snap exactly one video, Space pauses, hold ArrowRight for 2x) across Reels, Shorts, and TikTok. High framerates help oneshotting websites with complicated scroll-anchored animation logic (e.g. scale.com). Because of this reason, hwatu is optimized for consuming short-form content with much less resources than what you would have needed with Chromium or Firefox. The demo video below shows why hwatu is an excellent alternative browser option for your system, especially for tiling WMs:

Every window shares the one warm daemon (~56 MB per extra window), suspends when unfocused, and crash-restores at its last URL. Honest gaps: no Widevine or passkeys in WebKitGTK, so keep a fallback bound for Netflix. Ready-made WM configs (hyprland, sway, niri), the full keybind table, and setup: docs/human.md.

Features

• Headless / background / focused as a per-window property, switchable live
• Human hand-off: hwatu focus <id> drops the live session into your tiling WM
• Pixel-diff scoring: match percent + diff regions + heatmap (diff)
• Animations as numbers: duration, easing, velocity (motion)
• Deterministic animation frames: pin all animations at time t (seek)
• Page state as JSON, tokens not pixels (snapshot)
• Real input events with structured errors (click / type / scroll / upload)
• JS errors, console output, failed requests (console)
• Push event subscriptions as JSON lines or MCP notifications (watch)
• One-call page assertions with polling (expect)
• CAPTCHA / anti-bot detection with structured wait/resume (challenge)
• MCP server, plain CLI, and a 1-line JSON socket protocol
• A real browser for humans: mainstream keybinds, media-correct video, native ad blocking, crash restore

Why not Playwright or chrome-devtools-mcp?

There are three ways to give an agent a browser, and two of them are bad at it: