Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add Ikalus1988/MisakaNet --skill browser-harnessgit clone --depth 1 https://github.com/Ikalus1988/MisakaNetWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/ikalus1988/misakanet/browser-harness)<a href="https://agentmods.dev/skills/ikalus1988/misakanet/browser-harness"><img src="https://agentmods.dev/badge/skills/ikalus1988/misakanet/browser-harness/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/ikalus1988/misakanet/browser-harness"><img src="https://agentmods.dev/badge/skills/ikalus1988/misakanet/browser-harness.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00039 | $0.01509 |
| Opus 5 | $0.00019 | $0.00754 |
| Sonnet 5 | $0.00008 | $0.00302 |
| Haiku 4.5 | $0.00004 | $0.00151 |
Grade C, and why
browser-harness scanned grade C with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
curl -fsSL https://browser-use.com/profile.sh | sh Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -fsSL https://browser-use.com/profile.sh | sh How it starts
The opening of the file, as written. The whole thing — 155 lines — stays where its author put it; the contents beside it link to each section on GitHub.
browser-harness
AI 直连 Chrome 的薄 CDP 层。给 AI 一个 WebSocket,就能控制真实浏览器,AI 自己写辅助代码适应任何网站。
核心原则
- AI 是用户,不是人 — 框架为 AI 设计,不是为人设计的 UI
- 缺失即创造 — 遇到不会的,AI 自己写
agent_helpers.py学会 - CDP 是接口 — 直接给模型 Chrome DevTools Protocol,不抽象掉
快速开始
安装
# clone 到持久目录
git clone https://github.com/browser-use/browser-harness ~/Developer/browser-harness
cd ~/Developer/browser-harness
# 创建虚拟环境
uv venv .venv && source .venv/bin/activate
uv pip install -e .
# 连接到 Chrome(方式一:已有的 Chrome 开了 remote-debugging)
BU_CDP_URL=http://127.0.0.1:9222 browser-harness -c 'print(page_info())'
# 方式二:让 harness 启动云浏览器(需 BROWSER_USE_API_KEY)
start_remote_daemon("work")
连接 Chrome
方式 1 — 已有 Chrome 开 remote debugging(推荐): Chrome 菜单 → chrome://inspect/#remote-debugging → 勾选 "Allow remote debugging"
BU_CDP_URL=http://127.0.0.1:9222 browser-harness -c 'print(page_info())'
方式 2 — 隔离 profile:
google-chrome --remote-debugging-port=9222 --user-data-dir=/tmp/chrome-harness
BU_CDP_URL=http://127.0.0.1:9222 browser-harness -c '...'
常用命令
browser-harness -c 'new_tab("https://example.com"); wait_for_load(); print(page_info())'
browser-harness -c 'capture_screenshot()'
browser-harness -c 'new_tab("https://example.com"); wait_for_load(); js("document.title")'
核心 API
| 函数 | 作用 |
|---|---|
new_tab(url) |
新标签页打开 URL(不干扰当前页面) |
goto_url(url) |
在当前标签页导航 |
wait_for_load() |
等待页面加载完成 |
page_info() |
返回 {url, title, w, h, sx, sy, pw, ph} |
capture_screenshot() |
截图,返回文件路径 /tmp/shot.png |
js(expr) |
在页面执行 JS 表达式 |
click_at_xy(x, y) |
坐标点击(穿透 iframe/shadow DOM) |
scroll(dx, dy) |
滚动页面 |
cdp(domain, method, **params) |
直调 CDP |
list_tabs() |
列出所有标签页 |
表单填写
# 方式:js() 直接操作 DOM
browser-harness -c 'js(\'document.querySelector("input[name=custname]").value = "测试用户"\')'
# 验证
browser-harness -c 'js("return document.querySelector(\'input[name=custname]\').value")'
CDP 直调
# 任何 helper 没覆盖的,用 cdp()
cdp("Runtime.evaluate", expression="1+1", returnByValue=True)
# → {'result': {'type': 'number', 'value': 2}}
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago First seen · 155 lines · 39 tokens per session scan C 4b1f156b560b
browser-harness is a skill published in the GitHub repository Ikalus1988/MisakaNet (466 stars, last pushed yesterday), licensed Apache-2.0. It adds 39 tokens to every session and 1,509 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
playwright
Use when the task requires capturing or automating a real browser from the terminal.
e2e-testing-patterns
Master end-to-end testing with Playwright and Cypress to build reliable test suites that catch bugs, improve confidence, and enable fast deployment. Use when implementing E2E tests, debugging flaky tests, or establishing testing standards.
agent-browser
Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.
playwright-cli
Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.
azure-messaging-webpubsub-java
Build real-time web applications with Azure Web PubSub SDK for Java. Use when implementing WebSocket-based messaging, live updates, chat applications, or server-to-client push notifications.
computer-use
Read and drive native desktop applications through the accessibility layer — list on-screen apps, snapshot one window as a numbered element tree, then click / type / set a value / scroll / drag / run a named action, by element index or by screen coordinates. Use for work in a desktop app rather than a web page. Full…