kuri

justrach活跃维护

用 Zig 写的浏览器自动化与移动端控制工具，给 AI Agent 提供省 token 的 CDP 快照、HAR 录制和原生 adb 客户端，统一搞定网页抓取和 iOS/Android 设备操作。

Browser automation, web crawling, and iOS + Android device control for AI agents. Zig-native, token-efficient CDP snapshots, HAR recording, native adb wire-protocol client, and a standalone fetcher.

v0.6.1发布于 2026-10-09 09:25 UTC368star21fork同步于 2026-11-08 15:03 UTC

浏览器自动化移动端控制网页抓取 CDP 快照 adb 客户端

对比 GitHub 官网 ai-agents android browser-automation chrome-cdp ios

怎么上手

由 README 提炼，执行前请确认

原文

curl -fsSL https://kuri.trilok.ai/download | sh

硬件要求按 README 说明

README 未说明显存要求；CDP 模式需本机安装 Chrome/Chromium，独立抓取模式无需浏览器；预编译包支持 macOS arm64/x86_64 与 Linux x86_64/arm64。

支持 Apple Silicon

下载：源码 zip·GitHub Releases（安装包）本站不镜像安装包

项目说明

由 README 自动翻译提炼

没人工点评？写下你的经验

用 Zig 写的浏览器自动化与设备控制工具，为 AI agent 提供低 token 的 CDP 快照和 HTTP 接口

用 Zig 写的浏览器自动化与移动端控制工具，给 AI Agent 提供省 token 的 CDP 快照、HAR 录制和原生 adb 客户端，统一搞定网页抓取和 iOS/Android 设备操作。

核心能力

用 @eN 紧凑引用格式输出无障碍树快照，降低 token 消耗
提供 135 个 HTTP 端点，覆盖导航、点击、截图、HAR 录制等
POST /batch 单次调用批量执行多条命令，减少往返
内置原生 adb 协议客户端，可控制 iOS 与 Android 设备
kuri-fetch 独立抓取网页，无需 Chrome 运行时

容易踩的坑

从源码构建需 Zig ≥ 0.16.0；CDP 模式需本机有 Chrome/Chromium，否则用 kuri-fetch 独立模式；默认监听 127.0.0.1:8080，可用 KURI_SECRET 开启鉴权。

内容由模型根据 README 生成，可能不完整或不准确，以官方文档为准。看原文 README

用户点评

登录用户的真实使用记录，可按版本和硬件筛选

共 0 条

还没有人点评。第一个说出真实体验的人最有价值。

README 原文

未翻译

在 GitHub 打开

Install

curl -fsSL https://kuri.trilok.ai/download | sh

macOS arm64/x86_64 and Linux x86_64/arm64. Single binary, no runtime deps.

Direct downloads: macOS arm64 · macOS x86_64 · Linux x86_64 · Linux arm64

Browser automation & web crawling for AI agents. Written in Zig. Zero Node.js.

CDP automation · A11y snapshots · HAR recording · Standalone fetcher · Interactive terminal browser · Agentic CLI · Security testing · iOS + Android device control

Quick Start · Benchmarks · kuri-agent · Security Testing · API · Skills · Changelog

Why teams switch to Kuri: current Apple Silicon ReleaseFast builds stay sub-2 MB per binary, and a fresh Google Flights rerun on 2026-04-23 measured 3,392 tokens for a full kuri-agent loop (go→snap→click→snap→eval). Cross-tool deltas should be rerun in the same environment before quoting a percentage.

Why Kuri Wins for Agents

Most browser tooling was built for QA engineers. Kuri is built for agent loops: read the page, keep token cost low, act on stable refs, and move on.

135 HTTP endpoints — full parity with agent-browser and browser-use, from React inspection to Core Web Vitals.
7-12% fewer tokens than agent-browser on real pages thanks to @eN ref format and zero-prefix rendering.
44x lighter observations with /page/state (48 tokens) vs full snapshot (2,124 tokens) for the same Google Flights page.
Batch execution — POST /batch sends N commands in one HTTP call, eliminating N-1 round-trips and N-1 LLM turns.
React-compatible — trusted CDP mouse events and per-character key events fire React 18/19 onClick and onChange.

Snapshot tokens: Google Flights `SIN → TPE`

Fresh rerun on 2026-05-24 in this workspace, measured with wc -c and chars/4 approximation.

Tool / Mode	Chars	~Tokens	Note
`kuri snap` (full)	8,499	2,124	All nodes + interactive refs
`kuri snap` (interactive only)	~3,000	~750	Best for agent loops
`kuri /page/state`	190	48	Lightweight observation (url, title, scroll%, counts)
agent-browser snap (estimated)	~9,183	~2,295	`[ref=e0]` format overhead

Token efficiency: kuri vs agent-browser

Page	kuri tokens	agent-browser tokens	Savings
example.com	40	35	-13% (trivial page, agent-browser skips root)
Hacker News	386	~440	12% fewer
Google Flights SIN→TPE	2,124	~2,295	7% fewer

The savings come from kuri's compact format:

@e0 refs (3 chars) vs [ref=e0] (9 chars)
No - prefix per line (saves 2 chars × line count)
Same indentation, same node filtering

Full workflow cost: `go → snap → click → snap → eval`

Tool	Tokens per cycle
kuri-agent	~3,400
With `/page/state` instead of second snap	~1,700
With `POST /batch` (all in one call)	~1,700 (same tokens, 1 HTTP call instead of 5)

kuri vs libretto

libretto (Playwright + Node) is the closest competitor on per-step token cost. Measured head-to-head on 2026-07-04 — same Chrome, same tab, real tiktoken o200k_base counts (full methodology and reproduction: benchmarks/libretto_comparison.md). The honest split:

Axis	Winner	Detail
Latency per call	kuri	4–117 ms vs 1,344–1,500 ms (13–376× faster — persistent server vs Node-per-command)
Snapshot tokens, typical page	kuri	simple 61 vs 151 (2.5×), article 265 vs 363 (1.37×) — tighter grammar
Snapshot tokens, large list	split	kuri default 4,424 vs 813 — kuri emits all 259 refs, libretto truncates by default. With `limit=5` kuri renders 555 tokens (1.46× under libretto), 34 refs + `… +45 more` markers
Trajectory (feed, 9 clicks)	kuri, barely	898 vs 939 tokens (`limit=5` base + diff loop vs exec loop) — parity-to-slight-edge; the morning's 5.1× loss was the untruncated base
Repeat runs	libretto	compiles trajectories to a Playwright script → 0-token replays; kuri re-pays the loop every run

What kuri gained from studying libretto (all shipped this release): a diff-first loop (take_snapshot_diff, ~38 tokens/step); an adaptive diff that falls back to a full snapshot with a ! page replaced header on navigation; identity-only removal lines; screenshots written to disk (path returned, bytes never enter context); get_page_state over MCP; and — after rewriting parseA11yNodes as a real DFS tree walk — opt-in list truncation (/snapshot?limit=N, one … +K more line per capped run), scoped re-capture (scope=@ref), and hierarchy indentation, also exposed as uid/limit on MCP take_snapshot. The 9-click feed trajectory that cost 44,285 tokens with naive full re-snapshots costs 898 with a truncated base + diffs — 49× cheaper, and past libretto's 939.

The older tables above use a chars/4 token approximation; the libretto comparison uses real tiktoken counts. Rerun cross-tool numbers in your own environment before quoting a percentage.

Binary size and memory

Measured on Apple M4 Pro, macOS 26.4.1. Current binaries were built with -Doptimize=ReleaseFast.

Binary	Current size
`kuri`	1,093,840 B (1.04 MiB)
`kuri-agent`	629,904 B (615 KiB)
`kuri-browse`	1,089,120 B (1.04 MiB)
`kuri-fetch`	2,063,488 B (1.97 MiB)

RSS stayed flat across the Zig 0.16 migration

Measured against the current v0.4.3 ReleaseFast build with /usr/bin/time -l.

Command	`v0.4.3` mean max RSS
`kuri-fetch --version`	~2.45 MiB
`kuri-browse --version`	~2.45 MiB
`kuri-fetch --quiet --dump markdown http://example.com/`	~9.17 MiB

The Problem

Every browser automation tool drags in Playwright (~300 MB), a Node.js runtime, and a cascade of npm dependencies. Your AI agent just wants to read a page, click a button, and move on. Kuri is a single Zig binary. Four modes, zero runtime:

kuri           →  CDP server (Chrome automation, a11y snapshots, HAR)
kuri-fetch     →  standalone fetcher (no Chrome, QuickJS for JS, ~2 MB)
kuri-browse    →  interactive terminal browser (navigate, follow links, search)
kuri-agent     →  agentic CLI (scriptable Chrome automation + security testing)

📦 Installation

One-line install (macOS / Linux)

curl -fsSL https://raw.githubusercontent.com/justrach/kuri/release-channel/stable/install.sh | sh

Detects your platform, downloads the right binary, installs to ~/.local/bin. Downloads come from Kuri's self-managed release-channel branch. macOS binaries are locally signed with a Developer ID certificate. GitHub Release assets mirror these same tarballs.

bun / npm

bun install -g kuri-agent
# or: npm install -g kuri-agent

Downloads the correct native binary for your platform at install time.

Release channel

Kuri's stable binaries live on the release-channel branch and are served directly from GitHub raw URLs.

Stable installer: https://raw.githubusercontent.com/justrach/kuri/release-channel/stable/install.sh
Stable manifest: https://raw.githubusercontent.com/justrach/kuri/release-channel/stable/latest.json
Branch view: https://github.com/justrach/kuri/tree/release-channel/stable
Direct download pattern: https://raw.githubusercontent.com/justrach/kuri/release-channel/stable/<version>/kuri-<version>-<target>.tar.gz

Manual

Download the tarball for your platform from the stable release manifest or from the GitHub Releases page and unpack it to your $PATH.

Stable install URL:

curl -fsSL https://raw.githubusercontent.com/justrach/kuri/release-channel/stable/install.sh | sh

The manifest includes exact asset URLs plus SHA-256 checksums for aarch64-linux, x86_64-linux, aarch64-macos, and x86_64-macos.

Platform support

Platform	Status
macOS (`aarch64`, `x86_64`)	Prebuilt binaries, signed + notarized
Linux (`aarch64`, `x86_64`)	Prebuilt binaries
Windows (`x86_64`)	Experimental — cross-compile only. `zig build -Dtarget=x86_64-windows-gnu` is CI-verified, but Chrome automation, daemonization, signal-based shutdown, HAR recording, and the file-backed auth store are all stubbed with `error.UnsupportedOnWindows` at runtime. Use WSL2 if you need the real feature set. Tracked at #153.

Kuri leans on POSIX primitives (fork, clock_gettime, raw sockets) in several spots, so a full native Windows port is real work. The compile-level baseline above lets the --version/--help paths and pure in-memory operations run; the gnarly bits (Chrome, sockets, daemonize) need real Win32 implementations before they come off the stub list. If you want to take any of those on, +1 #153 or open a PR.

Build from source

Requires Zig ≥ 0.16.0.

git clone https://github.com/justrach/kuri.git
cd kuri
zig build -Doptimize=ReleaseFast
# Binaries in zig-out/bin/: kuri  kuri-agent  kuri-fetch  kuri-browse

⚡ Quick Start

Requirements: Zig ≥ 0.16.0 · Chrome/Chromium (for CDP mode)

git clone https://github.com/justrach/kuri.git
cd kuri

zig build              # build everything
zig build test         # run 252+ tests

# CDP mode — launches Chrome automatically
./zig-out/bin/kuri

# Standalone mode — no Chrome needed
./zig-out/bin/kuri-fetch https://example.com

# Interactive browser — browse from your terminal
./zig-out/bin/kuri-browse https://example.com

# Experimental standalone browser runtime — separate build, not production
(cd kuri-browser && zig build run -- render https://example.com)
(cd kuri-browser && zig build run -- bench --offline)

First run, shortest path

# start the server; if CDP_URL is unset, kuri launches managed Chrome for you
./zig-out/bin/kuri

# discover tabs from that managed browser
curl -s http://127.0.0.1:8080/discover

# inspect the discovered tab list
curl -s http://127.0.0.1:8080/tabs

Session-first agent loop

For agent-style HTTP usage, prefer a session header plus /tab/new, /page/info, and /snapshot instead of repeating tab_id on every call.

SESSION=hn-demo
BASE=http://127.0.0.1:8080

curl -s -H "X-Kuri-Session: $SESSION" \
  "$BASE/tab/new?url=https%3A%2F%2Fnews.ycombinator.com"

curl -s -H "X-Kuri-Session: $SESSION" "$BASE/page/info"
SNAP=$(curl -s -H "X-Kuri-Session: $SESSION" "$BASE/snapshot?filter=interactive&format=compact")
MORE_REF=$(printf '%s' "$SNAP" | python3 -c 'import re,sys; print(re.search(r"\"More\" @(e\\d+)", sys.stdin.read()).group(1))')
curl -s -H "X-Kuri-Session: $SESSION" "$BASE/action?action=click&ref=$MORE_REF"
curl -s -H "X-Kuri-Session: $SESSION" "$BASE/page/info"

There is also a thin experimental wrapper at tools/kuri_harness.py if you want Python helpers on top of the same HTTP surface.

If you already have Chrome running with remote debugging, set CDP_URL to either the WebSocket or HTTP endpoint:

CDP_URL=ws://127.0.0.1:9222/devtools/browser/... ./zig-out/bin/kuri
# or
CDP_URL=http://127.0.0.1:9222 ./zig-out/bin/kuri

Browse vercel.com in 4 commands

# 1. Discover Chrome tabs
curl -s http://localhost:8080/discover
# → {"discovered":1,"total_tabs":1}

# 2. Get tab ID
curl -s http://localhost:8080/tabs
# → [{"id":"ABC123","url":"chrome://newtab/","title":"New Tab"}]

# 3. Navigate
curl -s "http://localhost:8080/navigate?tab_id=ABC123&url=https://vercel.com"

# 4. Get accessibility snapshot (token-optimized for LLMs)
curl -s "http://localhost:8080/snapshot?tab_id=ABC123&filter=interactive"
# → [{"ref":"e0","role":"link","name":"VercelLogotype"},
#    {"ref":"e1","role":"button","name":"Ask AI"}, ...]

🌐 HTTP API

All endpoints return JSON. Optional auth via KURI_SECRET env var. 135 endpoints — full parity with agent-browser and browser-use.

Core

Path	Description
`GET /health`	Server status, tab count, version
`GET /tabs`	List all registered tabs
`GET /discover`	Auto-discover Chrome tabs via CDP
`GET /tab/current`	Get or set the current tab for an `X-Kuri-Session`
`GET /page/info`	Live URL/title/ready-state/viewport/scroll for the active tab
`GET /page/state`	Compact page observation: url, title, scroll%, viewport, counts of forms/links/images/inputs
`POST /batch`	Execute multiple commands in one HTTP call — returns array of results
`GET /browdie`	🌰 (easter egg)

Browser Control

Path	Params	Description
`GET /navigate`	`tab_id`, `url`	Navigate tab to URL
`GET /tab/new`	`url`, `activate`, `wait`	Create a new tab and optionally hydrate/set current tab
`GET /tab/close`	`tab_id`	Close a tab
`GET /window/new`	`url`, `activate`, `wait`	Create a new window/tab target
`GET /snapshot`	`tab_id`, `filter`, `format`	A11y tree snapshot with `eN` refs. Use `filter=interactive&format=compact` for low-token agent loops.
`GET /text`	`tab_id`	Extract page text
`GET /screenshot`	`tab_id`, `format`, `quality`, `save`	Capture screenshot (base64); `save=true` writes the PNG to `STATE_DIR/screenshots` and returns `{path,bytes}` instead
`GET /screenshot/annotated`	`tab_id`	Screenshot with numbered element labels
`GET /screenshot/diff`	`tab_id`, `baseline`	Visual diff between current and baseline screenshots
`GET /action`	`tab_id`, `ref`, `action`, `value`	Click/type/fill/select/scroll/hover/dblclick/check/uncheck/blur by ref
`GET /evaluate`	`tab_id`, `expression`	Execute JavaScript
`GET /evalhandle`	`tab_id`, `expression`	Execute JS, return objectId handle (not value)
`GET /close`	`tab_id`	Close tab + cleanup
`GET /bringtofront`	`tab_id`	Bring tab to front

Actions

Path	Params	Description
`GET /clear`	`ref`	Clear input field value
`GET /selectall`	`ref`	Select all text in input/contenteditable
`GET /setvalue`	`ref`, `value`	Set input value directly (bypasses key events)
`GET /dispatch`	`ref`, `type`	Dispatch custom DOM event on element
`GET /boundingbox`	`ref`	Get element bounding rect (x, y, width, height, centerX, centerY)
`GET /getattribute`	`ref`, `name`	Get element attribute by name
`GET /inputvalue`	`ref`	Get current input element value
`GET /element/state`	`ref`, `check`	Quick boolean: `exists`, `visible`, `enabled`, `checked`
`GET /find-element`	`text`/`role`/`label`/`placeholder`/`testid`	Semantic locator — find element without snap
`GET /highlight`	`ref` or `selector`	Highlight element with overlay

Mouse & Touch

Path	Params	Description
`GET /mouse/move`	`x`, `y`	Move mouse to coordinates
`GET /mouse/down`	`x`, `y`, `button`	Mouse button down
`GET /mouse/up`	`x`, `y`, `button`	Mouse button up
`GET /mouse/wheel`	`x`, `y`, `deltaX`, `deltaY`	Mouse wheel scroll
`GET /tap`	`x`, `y`	Touch tap (touchStart + touchEnd)
`GET /swipe`	`startX`, `startY`, `endX`, `endY`	Touch swipe gesture
`GET /drag`	`src_ref`, `tgt_ref`	Drag element to target

Keyboard

Path	Params	Description
`GET /keyboard/type`	`tab_id`, `text`	Type text via key events
`GET /keyboard/inserttext`	`tab_id`, `text`	Insert text directly
`GET /keydown`	`tab_id`, `key`	Key down event
`GET /keyup`	`tab_id`, `key`	Key up event

Content Extraction

Path	Description
`GET /markdown`	Convert page to Markdown
`GET /links`	Extract all links
`GET /dom/query`	CSS selector query
`GET /dom/html`	Get element HTML
`GET /dom/attributes`	Get element attributes
`GET /pdf`	Print page to PDF
`GET /find`	In-page text search

Waiting

Path	Params	Description
`GET /wait`	`selector`, `text`, `url`, `state`, `visible`, `timeout`	Wait for selector/text/URL pattern/networkidle/load state
`GET /wait/function`	`expression`, `timeout`	Wait for arbitrary JS expression to be truthy
`GET /wait/download`	`timeout`	Wait for file download to complete

Dialog Handling

Path	Description
`GET /dialog/auto`	Auto-handle all JS dialogs (accept or dismiss)
`GET /dialog/accept`	Accept current dialog (with optional prompt text)
`GET /dialog/dismiss`	Dismiss current dialog

Network & HAR

Path	Description
`GET /har/start`	Start recording network traffic
`GET /har/stop`	Stop + return HAR 1.2 JSON
`GET /har/status`	Recording state + entry count
`GET /har/replay`	API map with curl/fetch/python code snippets
`GET /cookies`	Get cookies
`GET /cookies/set`	Set cookies
`GET /cookies/delete`	Delete cookies
`GET /cookies/clear`	Clear all cookies
`GET /headers`	Set custom request headers
`GET /intercept/start`	Start request interception
`GET /intercept/stop`	Stop request interception
`GET /intercept/requests`	List intercepted requests
`GET /request/detail`	Get response body for a request ID
`GET /response/body`	Fetch URL and return response body
`GET /network`	Network traffic stats
`GET /download`	Trigger file download

Navigation & State

Path	Description
`GET /back`	Browser back
`GET /forward`	Browser forward
`GET /reload`	Reload page
`GET /stop`	Stop page loading
`GET /pushstate`	SPA navigation via history.pushState
`GET /storage/local`	Get/set localStorage
`GET /storage/session`	Get/set sessionStorage
`GET /storage/local/clear`	Clear localStorage
`GET /storage/session/clear`	Clear sessionStorage
`GET /session/save`	Save browser session
`GET /session/load`	Restore browser session
`GET /session/list`	List saved sessions
`GET /setcontent`	Set page HTML directly (POST)

Auth Profiles

Path	Description
`GET /auth/profile/save`	Save cookies + storage as a named auth profile
`GET /auth/profile/load`	Restore a named auth profile into a tab
`GET /auth/profile/list`	List saved auth profiles
`GET /auth/profile/delete`	Delete a saved auth profile
`GET /auth/extract`	Extract auth tokens (JWT, cookies, headers)
`GET /set/credentials`	Set HTTP basic auth credentials

On macOS, auth profile secrets are stored in the user Keychain.

Emulation

Path	Params	Description
`GET /emulate`	device type, screen size	Device emulation
`GET /set/viewport`	`width`, `height`	Set viewport size
`GET /set/useragent`	`ua`	Set user agent
`GET /set/media`	`media`	Emulate media type
`GET /set/offline`	`offline`	Toggle offline mode
`GET /geolocation`	`lat`, `lng`	Override geolocation
`GET /timezone`	`timezone`	Override timezone (e.g. `America/New_York`)
`GET /locale`	`locale`	Override locale (e.g. `en-US`)
`GET /permissions`	`name`, `state`	Grant/deny permissions (geolocation, notifications, clipboard)

Scripts & Injection

Path	Description
`GET /script/inject`	Inject JavaScript into page (persists across nav)
`GET /initscript/remove`	Remove a previously injected init script
`GET /addstyle`	Inject CSS stylesheet
`GET /expose`	Expose a named function to page JS context

React Inspection

Path	Description
`GET /react/tree`	React component tree via DevTools hook
`GET /react/inspect`	React component props and state
`GET /react/renders`	React render tracking (start/stop)
`GET /react/suspense`	React Suspense boundary status

Recording & Performance

Path	Description
`GET /recording/start`	Record user actions (click, input, navigate)
`GET /recording/stop`	Stop recording + return action log
`GET /vitals`	Core Web Vitals (LCP, CLS, FID, TTFB, FCP, domInteractive)
`GET /perf/lcp`	Largest Contentful Paint timing
`GET /trace/start`	Start performance trace
`GET /trace/stop`	Stop trace
`GET /profiler/start`	Start JS profiler
`GET /profiler/stop`	Stop profiler

Debugging

Path	Description
`GET /debug/enable`	Enable in-page debug HUD and optional freeze mode
`GET /debug/disable`	Disable in-page debug HUD
`GET /inspect`	Element inspection
`GET /errors`	Collect JS errors
`GET /console`	Read console logs
`GET /frames`	List page frames
`GET /frame`	Switch to iframe context by name or URL
`GET /mainframe`	Switch back to main frame
`GET /diff/snapshot`	Compact `+`/`~`/`-` diff vs the previous call for this tab — the token-efficient action loop (aliased by `/snapshot/changes`). Falls back to a full snapshot with a `! page replaced` header on mass change.
`GET /diff/url`	Compare two URLs side by side (navigate, snapshot, diff)

Streaming

Path	Description
`GET /screencast/start`	Start screen recording
`GET /screencast/stop`	Stop screen recording
`GET /video/start`	Start video capture
`GET /video/stop`	Stop video capture
`GET /ws/start`	WebSocket tunnel start
`GET /ws/stop`	WebSocket tunnel stop

Agent-friendly loop

The lowest-friction server loop is:

GET /tab/new?url=...
GET /page/state (lightweight) or GET /snapshot?filter=interactive&format=compact (full)
GET /action?action=click&ref=eN
Repeat — or use POST /batch for multi-step operations in one call

`url` and `expression` query params are percent-decoded. Send `X-Kuri-Session: my-agent` to persist tab context server-side.

🧠 Skills

The repo includes a user-extensible skill area:

skills/kuri-skill.md is the base Kuri HTTP-agent skill
skills/custom/ is reserved for your own project-specific skills
skills/custom/hackernews-page-2.md is a concrete example custom skill
.claude/skills/kuri-server/SKILL.md stays in sync for Claude-style repo skills

The base skill now also explains which browser path to use:

kuri HTTP API: production Chrome/CDP automation with sessions, snapshots, actions, HAR, cookies, and screenshots
kuri-fetch: standalone no-Chrome fetch/text extraction
kuri-browse: interactive terminal browsing
kuri-agent: scriptable CLI automation against the Kuri server
kuri-browser/: experimental separate Zig-native browser runtime for parity work

For the experimental browser CLI:

cd kuri-browser
zig build run -- render https://news.ycombinator.com --selector ".titleline a" --dump text
zig build run -- render https://todomvc.com/examples/react/dist/ --js --wait-eval "document.querySelectorAll('.todo-list li').length >= 1"
zig build run -- parity --offline
zig build run -- bench --offline
zig build run -- serve-cdp --port 9333

kuri-browser serve-cdp exposes Chrome-style HTTP discovery plus a minimal WebSocket JSON-RPC router for protocol smoke tests. Runtime eval returns V8-shaped CDP remote objects backed by QuickJS; this does not add a V8 dependency and is not full Playwright/Puppeteer compatibility yet.

Screenshots in kuri-browser currently delegate to the main Kuri/CDP renderer. Start ./zig-out/bin/kuri first, then:

cd kuri-browser
zig build run -- screenshot https://example.com --out example.jpg --compress --kuri-base http://127.0.0.1:8080

--compress captures a PNG baseline and JPEG candidate, writes the smaller file, and reports byte savings. Current local measurement on https://example.com: 20,523 bytes PNG to 18,183 bytes JPEG quality 50, saving 2,340 bytes or 11%.

Advanced

Path	Description
`GET /diff/snapshot`	Compact `+`/`~`/`-` delta vs the previous snapshot (agent action loop)
`GET /emulate`	Device emulation
`GET /geolocation`	Set geolocation
`POST /upload`	File upload
`GET /script/inject`	Inject JavaScript
`GET /intercept/start`	Start request interception
`GET /intercept/stop`	Stop interception
`GET /screenshot/annotated`	Screenshot with element annotations
`GET /screenshot/diff`	Visual diff between screenshots
`GET /screencast/start`	Start screencast
`GET /screencast/stop`	Stop screencast
`GET /video/start`	Start video recording
`GET /video/stop`	Stop video recording
`GET /console`	Get console messages
`GET /stop`	Stop page loading
`GET /get`	Direct HTTP fetch (server-side)
`GET /scrollintoview`	Scroll a referenced element into view
`GET /drag`	Drag from one ref to another
`GET /keyboard/type`	Type text with key events
`GET /keyboard/inserttext`	Insert text directly
`GET /keydown`	Dispatch a keydown event
`GET /keyup`	Dispatch a keyup event
`GET /wait`	Wait for ready state or element conditions
`GET /tab/close`	Close a tab
`GET /highlight`	Highlight an element by ref or selector
`GET /errors`	Get page/runtime errors
`GET /set/offline`	Toggle offline network emulation
`GET /set/media`	Set emulated media features
`GET /set/credentials`	Set HTTP basic auth credentials
`GET /find`	Find text matches in the current page
`GET /trace/start`	Start Chrome tracing
`GET /trace/stop`	Stop tracing and return trace data
`GET /profiler/start`	Start JS profiler
`GET /profiler/stop`	Stop JS profiler
`GET /inspect`	Inspect an element or page state
`GET /set/viewport`	Set viewport size
`GET /set/useragent`	Override user agent
`GET /dom/attributes`	Get element attributes
`GET /frames`	List frame tree
`GET /network`	Inspect network state/requests

🛡️ Stealth & Bot Evasion

Kuri applies anti-detection patches automatically on startup — no manual config needed.

What's applied

Page.addScriptToEvaluateOnNewDocument — stealth patches run before any page JS
navigator.webdriver = false — hides automation flag at Chromium level (--disable-blink-features=AutomationControlled)
WebGL/Canvas/AudioContext spoofing — defeats fingerprint-based detection
UA rotation — 5 realistic Chrome/Safari/Firefox user agents
chrome.csi/chrome.loadTimes — stubs for Akamai-specific checks

Bot block detection

Navigate auto-detects blocks and returns structured fallback:

curl -s "http://localhost:8080/navigate?tab_id=ABC&url=https://protected-site.com"
# If blocked:
# {"blocked":true,"blocker":"akamai","ref_code":"0.7d...",
#  "fallback":{"suggestions":["Open URL directly in browser","Use KURI_PROXY"]}}
# If ok: normal CDP response

Detects: Akamai, Cloudflare, PerimeterX, DataDome, generic captcha.

Proxy support

KURI_PROXY=socks5://user:pass@residential-proxy:1080 ./zig-out/bin/kuri
KURI_PROXY=http://proxy:8080 ./zig-out/bin/kuri

Tested sites

Site	Protection	Result
Singapore Airlines	Akamai WAF	✅ Bypassed (was blocked before v0.4.0)
Shopee SG	Custom anti-fraud	✅ Page loads, redirects to login
Google Flights	None	✅ Full interaction
Booking.com	PerimeterX	⚠️ Needs proxy

🔧 kuri-fetch

Standalone HTTP fetcher — no Chrome, no Playwright, no npm. Ships as a ~2 MB binary with built-in QuickJS for JS execution.

zig build fetch    # build + run

# Default: convert to Markdown
kuri-fetch https://example.com

# Extract links
kuri-fetch -d links https://news.ycombinator.com

# Structured JSON output
kuri-fetch --json https://example.com

# Execute inline scripts via QuickJS
kuri-fetch --js https://example.com

# Write to file, quiet mode
kuri-fetch -o page.md -q https://example.com

# Pipe-friendly: content → stdout, status → stderr
kuri-fetch -d text https://example.com | wc -w

Features

5 output modes — markdown, html, links, text, json
QuickJS JS engine — --js executes inline <script> tags
DOM stubs — document.querySelector, getElementById, window.location, document.title, console.log, setTimeout (SSR-style)
SSRF defense — blocks private IPs, metadata endpoints, non-HTTP schemes
Colored output — respects NO_COLOR, TERM=dumb, --no-color, TTY detection
File output — -o / --output with byte count + timing summary
Custom UA — --user-agent flag
Quiet mode — -q suppresses stderr status

🌐 kuri-browse

Interactive terminal browser — browse the web from your terminal. No Chrome needed.

zig build browse   # build + run

kuri-browse https://example.com

🌰 kuri-browse — terminal browser
→ loading https://example.com

# Example Domain
This domain is for use in documentation examples...
Learn more [1]

───── Links ─────
  [1] https://iana.org/domains/example

✓ 528 bytes, 1 links (133ms)
[nav] https://example.com> 1     ← type 1 to follow the link

Commands

Command	Action
`<number>`	Follow link [N]
`<url>`	Navigate (if contains `.`)
`:go <url>`	Navigate to URL
`:back`, `:b`	Go back in history
`:forward`, `:f`	Go forward
`:reload`, `:r`	Re-fetch current page
`:links`, `:l`	Show link index
`/<term>`	Search in page (highlights matches)
`:search <t>`	Search in page
`:n`, `:next`	Re-highlight search
`:history`	Show navigation history
`:help`, `:h`	Show all commands
`:quit`, `:q`	Exit

Features

Colored markdown rendering — headings, links, code blocks, bold, blockquotes
Numbered links — every link gets [N], type the number to follow it
Navigation history — back/forward like a real browser
In-page search — /term highlights all matches
Relative URL resolution — follows links naturally across pages
Smart filtering — skips javascript: and mailto: hrefs

🤖 kuri-agent

Scriptable CLI for Chrome automation — drives the browser command-by-command from your terminal or shell scripts. Shares session state across invocations via ~/.kuri/session.json.

zig build agent   # build kuri-agent

# 1. Find a Chrome tab
kuri-agent tabs
# → ws://127.0.0.1:9222/devtools/page/ABC123  https://example.com

# 2. Attach to it
kuri-agent use ws://127.0.0.1:9222/devtools/page/ABC123

# 3. Navigate + interact
kuri-agent go https://example.com
kuri-agent snap --interactive        # → [{"ref":"e0","role":"link","name":"More info"}]
kuri-agent click e0
kuri-agent shot                      # saves ~/.kuri/screenshots/<ts>.png

Commands

Command	Description
`tabs [--port N]`	List Chrome tabs
`use <ws_url>`	Attach to a tab (saves session)
`open [url] [--port N]`	Open a new tab (optionally navigating to url)
`status`	Show current session
`go <url>`	Navigate to URL
`snap [--interactive] [--json] [--text] [--depth N]`	A11y snapshot, saves `eN` refs
`click <ref>`	Click element by ref (CDP mouse events, React-compatible)
`type <ref> <text>`	Type into element (per-character key events, React-compatible)
`fill <ref> <text>`	Fill input value
`select <ref> <value>`	Select dropdown option
`hover <ref>`	Hover over element
`focus <ref>`	Focus element
`scroll`	Scroll the page
`viewport [width height]`	Get or set viewport dimensions
`eval <js>`	Evaluate JavaScript
`text [selector]`	Get page text
`shot [--out file.png]`	Screenshot
`back`	Navigate back
`forward`	Navigate forward
`reload`	Reload current page
`cookies`	List cookies with security flags
`headers`	Check security response headers
`audit`	Full security audit
`storage [local\|session\|all]`	Dump localStorage / sessionStorage
`jwt`	Extract and decode JWTs from cookies and storage
`fetch <method> <url> [--data <json>]`	Authenticated fetch using page cookies
`probe <url-template> <start> <end>`	IDOR probe: iterate numeric IDs in URL
`grab <ref>`	Click ref, intercept `window.open`, follow redirect in-tab
`wait-for-tab [--port N]`	Poll for a new tab, auto-switch session
`stealth`	Apply anti-detection patches
`set-header <name> <value>`	Add a custom header to all requests
`show-headers`	Show stored extra headers
`clear-headers`	Remove all extra headers

📱 kuri-mobile (iOS + Android)

Native Zig CLI for driving iOS Simulators, real iPhones (listing + launch/terminate), and Android devices/emulators — inspired by mobile-device-mcp, reimplemented in Zig with no Bun/Node/Gradle/Xcode in the build path.

cd kuri-mobile && zig build && cp zig-out/bin/kuri-mobile ../zig-out/bin/

# The main `kuri` binary forwards android/ios subcommands to kuri-mobile:
kuri ios list-devices                              # sims + real devices (usbmuxd, native)
kuri ios openurl https://example.com               # navigate Safari
kuri ios screenshot out.png                        # auto-picks booted sim
kuri ios launch com.apple.Preferences

kuri android list-devices                          # native Zig adb wire-protocol client
kuri android tap 540 1200
kuri android swipe 100 1500 100 500
kuri android screenshot phone.png
kuri android uitree                                # flat element list via uiautomator dump

What's native Zig: adb host protocol (libc sockets, 4-hex framing over host:transport:/shell:/exec:), Android XML UI tree parser, usbmuxd ListDevices plist client. What shells out: xcrun simctl (iOS Simulator), xcrun devicectl (iOS real-device launch/terminate). Driverless by design: no on-device app is installed, so run_code sandboxes and XCUITest-backed tap/uitree on real iOS devices are intentionally not available. See kuri-mobile/README.md for the full parity matrix vs upstream.

🔒 Security Testing

kuri-agent supports browser-native security trajectories — log in once, then run reconnaissance and header/cookie audits without leaving the terminal.

Trajectories

Enumerate → Inspect — after authenticating, dump auth cookies and check security flags:

kuri-agent go https://target.example.com/login
kuri-agent snap --interactive
kuri-agent fill e2 myuser
kuri-agent fill e3 mypassword
kuri-agent click e4                  # submit login

kuri-agent cookies
# cookies (3):
#   session_id  domain=.example.com path=/  [Secure] [HttpOnly] [SameSite=Strict]
#   csrf_token  domain=.example.com path=/  [Secure] [!HttpOnly]
#   tracking    domain=.example.com path=/  [!Secure] [!HttpOnly]

Header audit — check what security headers the target sends:

kuri-agent go https://target.example.com
kuri-agent headers
# → {"url":"https://...","status":200,"headers":{
#     "content-security-policy":"default-src 'self'",
#     "strict-transport-security":"max-age=31536000",
#     "x-frame-options":"(missing)",
#     "x-content-type-options":"nosniff", ...}}

Full audit — HTTPS, missing headers, JS-visible cookies in one shot:

kuri-agent audit
# → {"protocol":"https:","url":"https://...","score":6,
#     "issues":["MISSING:x-frame-options","COOKIES_EXPOSED_TO_JS:2"],
#     "headers":{"content-security-policy":"default-src 'self'", ...}}

Cross-account trajectory — use eval to replay API calls with different tokens:

# After login, grab the auth token from localStorage
kuri-agent eval "localStorage.getItem('token')"

# Probe a resource ID with the current session
kuri-agent eval "fetch('/api/assessments/42').then(r=>r.status)"

# Check for IDOR: does a different user's resource return 200 or 403?
kuri-agent eval "fetch('/api/assessments/99').then(r=>r.status)"

Trajectory Report Format

kuri-agent outputs JSON suitable for pipeline integration. Each security command emits a single JSON line — pipe through jq for triage:

kuri-agent audit | jq '.issues[]'
kuri-agent cookies | head -20
kuri-agent headers | jq '.headers | to_entries[] | select(.value == "(missing)") | .key'

🏗 Architecture

┌──────────────────────────────────────────────────────────┐
│                     HTTP API Layer                        │
│         (std.http.Server, thread-per-connection)          │
├──────────────┬──────────────────┬────────────────────────┤
│   Browser    │  Crawler Engine  │   kuri-fetch / browse   │
│   Bridge     │                  │   (standalone CLIs)     │
├──────────────┼──────────────────┼────────────────────────┤
│ CDP Client   │ URL Validator    │ std.http.Client         │
│ Tab Registry │ HTML→Markdown    │ QuickJS JS Engine       │
│ A11y Snapshot│ Link Extractor   │ DOM Stubs (Layer 3)     │
│ Ref Cache    │ Text Extractor   │ SSRF Validator          │
│ HAR Recorder │                  │ Colored Renderer        │
│ Stealth JS   │                  │ History + REPL          │
├──────────────┴──────────────────┴────────────────────────┤
│  Chrome Lifecycle Manager                                 │
│  (launch, health-check, auto-restart, port detection)     │
└──────────────────────────────────────────────────────────┘

Memory Model

Arena-per-request — all per-request memory freed in one deinit() call
No GC — GeneralPurposeAllocator in debug mode catches every leak
Proper cleanup chains — Launcher → Bridge → CdpClients → HarRecorders → Snapshots → Tabs
errdefer guards — partial failures roll back cleanly

Chrome Lifecycle

Mode	Behavior
Managed (no `CDP_URL`)	Launches Chrome headless, finds free CDP port, supervises, auto-restarts on crash (max 3 retries), kills on shutdown
External (`CDP_URL` set)	Connects to existing Chrome, health-checks via `/json/version`, does NOT kill on shutdown

📁 Structure

kuri/
├── build.zig                  # Build system (Zig 0.16.0)
├── build.zig.zon              # Package manifest + QuickJS dep
├── src/
│   ├── main.zig               # CDP server entry point
│   ├── fetch_main.zig         # kuri-fetch CLI entry point
│   ├── browse_main.zig        # kuri-browse CLI entry point
│   ├── js_engine.zig          # QuickJS wrapper + DOM stubs
│   ├── bench.zig              # Benchmark harness
│   ├── chrome/
│   │   └── launcher.zig       # Chrome lifecycle manager
│   ├── server/
│   │   ├── router.zig         # HTTP route dispatch (40+ endpoints)
│   │   ├── middleware.zig     # Auth (constant-time comparison)
│   │   └── response.zig      # JSON response helpers
│   ├── bridge/
│   │   ├── bridge.zig         # Central state (tabs, CDP, HAR, snapshots)
│   │   └── config.zig         # Env var configuration
│   ├── cdp/
│   │   ├── client.zig         # CDP WebSocket client
│   │   ├── websocket.zig      # WebSocket frame codec
│   │   ├── protocol.zig       # CDP method constants
│   │   ├── actions.zig        # High-level CDP actions
│   │   ├── stealth.zig        # Bot detection bypass
│   │   └── har.zig            # HAR 1.2 recorder
│   ├── snapshot/
│   │   ├── a11y.zig           # A11y tree with interactive filter
│   │   ├── diff.zig           # Snapshot delta diffing
│   │   └── ref_cache.zig      # eN ref → node ID cache
│   ├── crawler/
│   │   ├── validator.zig      # SSRF defense, URL validation
│   │   ├── markdown.zig       # HTML → Markdown (SIMD tag counting)
│   │   ├── fetcher.zig        # Page fetching
│   │   ├── extractor.zig      # Readability extraction
│   │   └── pipeline.zig       # Parallel crawl pipeline
│   ├── storage/
│   │   ├── local.zig          # Local file writer
│   │   └── r2.zig             # R2/S3 uploader
│   ├── util/
│   │   └── json.zig           # JSON helpers
│   └── test/
│       ├── harness.zig        # Test HTTP client
│       ├── integration.zig    # Integration tests
│       └── merjs_e2e.zig      # E2E tests
├── js/
│   ├── stealth.js             # Bot detection bypass
│   └── readability.js         # Content extraction
├── kuri-browser/              # Native Zig rendering experiments
└── kuri-mobile/               # iOS + Android device control (Zig-native adb + usbmuxd)
    ├── src/
    │   ├── common/            # io helpers, unified UI tree parser
    │   ├── android/           # adb wire protocol client, driver, CLI
    │   └── ios/               # simctl, usbmuxd, devicectl, CLI
    └── README.md              # Full parity matrix vs mobile-device-mcp

⚙️ Configuration

Env Var	Default	Description
`HOST`	`127.0.0.1`	Server bind address
`PORT`	`8080`	Server port
`CDP_URL`	(none)	Connect to existing Chrome (`ws://...` or `http://127.0.0.1:9222`)
`KURI_SECRET`	(none)	Auth secret for API requests
`STATE_DIR`	`.kuri`	Session state directory
`REQUEST_TIMEOUT_MS`	`30000`	HTTP request timeout
`NAVIGATE_TIMEOUT_MS`	`30000`	Navigation timeout
`STALE_TAB_INTERVAL_S`	`30`	Stale tab cleanup interval
`NO_COLOR`	(none)	Disable colored CLI output

💰 Token Cost

For a 50-page monitoring task (from Pinchtab benchmarks):

Method	Tokens	Cost ($)	Best For
`/text`	~40,000	$0.20	Read-heavy (13× cheaper than screenshots)
`/snapshot?filter=interactive&format=compact`	~40,000	$0.20	Low-token element interaction
`/snapshot` (full)	~525,000	$2.63	Full page understanding
`/screenshot`	~100,000	$1.00	Visual verification

🤝 Contributing

Open an issue before submitting a large PR so we can align on the approach.

git clone https://github.com/justrach/kuri.git
cd kuri
zig build test         # 252+ tests must pass
zig build test-fetch   # kuri-fetch tests (69 tests)
zig build test-browse  # kuri-browse tests (22 tests)

See CONTRIBUTORS.md for guidelines.

Credits

Project	What we borrowed
agent-browser	`@eN` ref system, snapshot diffing, HAR recording patterns
Pinchtab	Browser control architecture for AI agents
Pathik	High-performance crawling patterns
QuickJS-ng via mitchellh/zig-quickjs-ng	JS engine for `kuri-fetch`
Lightpanda	Zig-native headless browser pioneer, CDP compatibility patterns
Zig 0.16.0	The whole stack

License

Apache-2.0

怎么上手

项目说明

核心能力

容易踩的坑

最近发版

v0.6.1

Fixes

v0.6.0

Features

Removed

v0.5.1

Features

Fixes

Release verification

v0.4.14

Fixes

Features — Android

Tests

v0.4.13

Fixes — routes that fabricated their answers

Fixes — script injection

Fixes — memory safety

用户点评

README 原文

Install

Why Kuri Wins for Agents

Snapshot tokens: Google Flights SIN → TPE

Token efficiency: kuri vs agent-browser

Full workflow cost: go → snap → click → snap → eval

kuri vs libretto

Binary size and memory

RSS stayed flat across the Zig 0.16 migration

The Problem

📦 Installation

One-line install (macOS / Linux)

bun / npm

Release channel

Manual

Platform support

Build from source

⚡ Quick Start

First run, shortest path

Session-first agent loop

Browse vercel.com in 4 commands

🌐 HTTP API

Core

Browser Control

Actions

Mouse & Touch

Keyboard

Content Extraction

Waiting

Dialog Handling

Network & HAR

Navigation & State

Auth Profiles

Emulation

Scripts & Injection

React Inspection

Recording & Performance

Debugging

Streaming

Agent-friendly loop

url and expression query params are percent-decoded. Send X-Kuri-Session: my-agent to persist tab context server-side.

🧠 Skills

Advanced

🛡️ Stealth & Bot Evasion

What's applied

Bot block detection

Proxy support

Tested sites

🔧 kuri-fetch

Features

🌐 kuri-browse

Commands

Features

🤖 kuri-agent

Commands

📱 kuri-mobile (iOS + Android)

🔒 Security Testing

Snapshot tokens: Google Flights `SIN → TPE`

Full workflow cost: `go → snap → click → snap → eval`

`url` and `expression` query params are percent-decoded. Send `X-Kuri-Session: my-agent` to persist tab context server-side.