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 agentmods add instructions/zzhan111/multi-agents-browser/claude-mdgit clone --depth 1 https://github.com/zzhan111/multi-agents-browserWrote 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/instructions/zzhan111/multi-agents-browser/claude-md)<a href="https://agentmods.dev/instructions/zzhan111/multi-agents-browser/claude-md"><img src="https://agentmods.dev/badge/instructions/zzhan111/multi-agents-browser/claude-md.svg" alt="Measured on agentmods" 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 | $0.02053 | $0.02053 |
| Opus 5 | $0.01026 | $0.01026 |
| Sonnet 5 | $0.00411 | $0.00411 |
| Haiku 4.5 | $0.00205 | $0.00205 |
Grade A, and why
multi-agents-browser CLAUDE.md scanned grade A with 0 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 3d 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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 137 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
项目改名 → MultiAgentsBrowser / ma-browser（2026-06-10）
本项目由 bb-browser（fork 自 epiral/bb-browser）改名为 MultiAgentsBrowser，CLI/MCP 名 ma-browser，迁到独立仓库 github.com/zzhan111/multi-agents-browser（保留 epiral credit，见 README/LICENSE）。本地工作目录仍是 Z:\Apps\bb-browser。
作为历史兼容层有意保留、改动代码时勿动：
BB_*环境变量（BB_BROWSER_HOME/BB_SESSION_*/BB_TRACE_CAPACITY等）~/.bb-browser数据目录、x-bb-*HTTP 头、bbTabId协议字段Z:\Apps\bb-browser-tray部署目录名（exe 已改ma-browser-tray.exe）epiral/bb-sites社区适配器仓库链接、CHANGELOG.md（带 legacy 注的历史条目）
新增改动应使用 ma-browser 命名。运维链路细节见用户记忆 project-rename-ma-browser / bb-chain-guard / deployment-after-commit。
Trace Export — 已完成改进（2026-05-18）
以下原 roadmap 项已全部实现，记录于此供历史回溯。当前如需新增改进，请在文末"后续待改进项"追加。
✅ P1: 导航事件录制
trace-inject.ts监听popstate和hashchange，捕获 SPA 内导航cdp-connection.ts在Page.frameNavigated（仅主框架）时发出navigationtrace 事件tab-state.ts的addTraceNavigation按 URL 去重，避免 CDP redirect 链产生重复- 导出脚本中渲染为
page.goto(url)；与activeTab.url相同的首条 navigation 会被跳过避免双重 goto
✅ P1: 选择器可移植性
ExportDialog 提供三个模式：
- Auto：
ref→cssSelector→xpath（保留 ma-browser ref 优先级） - CSS：
cssSelector→xpath，便于在标准 Playwright/Selenium 环境运行 - XPath：仅用
xpath，跨 DOM 变化更稳定
JS/Playwright 在 XPath 模式下使用 xpath=... 前缀；Python 切换到 By.XPATH。
✅ P2: 步骤间等待机制
ExportDialog 新增"智能等待"开关（默认开启）：
- JS / Playwright：在 click/fill/select/check 前插入
page.waitForSelector(sel)或locator(sel).waitFor() - Python：插入
wait.until(EC.presence_of_element_located((by, sel)))
✅ P2: 字符串转义
esc() 替换为 q(str) → 返回 JSON.stringify(str)，自动处理换行、制表符、Unicode、引号、反斜杠。生成的字符串字面量从单引号迁移到双引号。
✅ P3: Ring Buffer 上限
TRACE_CAPACITY 通过 BB_TRACE_CAPACITY 环境变量配置（默认 1000，最小 100）。缓冲首次写满时打印一次性 warning 到 daemon stderr：
[ma-browser] trace buffer full for tab <shortId> (cap=1000); oldest events will be discarded. Set BB_TRACE_CAPACITY to raise the limit.
后续待改进项
（暂无。）
线 C — 状态持久面 (State Persistence Plane) 实施记录（2026-06-03）
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.
- 3d ago First seen · 137 lines · 2,053 tokens per session scan A ffdf8dc3655f
multi-agents-browser CLAUDE.md is an instructions file published in the GitHub repository zzhan111/multi-agents-browser (1 stars, last pushed today), licensed MIT. It adds 2,053 tokens to every session, about $0.0103 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
Browser4 CLAUDE.md
Claude Code instructions for platonai/Browser4, covering browser4 — project context for claude, architecture, key dispatch chain (cli → browser), batch commands and e2e test structure.
fix-quera AGENTS.md
Instructions for AlirezaKeshavarz83/fix-quera, covering agent guidance, project shape, visual style, quera page data findings and compatibility findings.
fast-browser CLAUDE.md
Claude Code instructions for m4ttstack/fast-browser, covering fast browser plugin, where a change belongs, fork branch: use fast-browser-runtime, releasing a new runtime and re-pinning this repo: use the script.
sendblue-browser-use AGENTS.md
Instructions for sendblue-api/sendblue-browser-use, covering agents.md — sendblue-browser-use, what this repo is, setup, common commands and health (no auth).
sniff AGENTS.md
AGENTS.md instructions for Aboudjem/sniff, covering agents.md: sniff, what this repo is, how an agent should use sniff, handling the playwright setup gate and finding output schema.
browser_oxide AGENTS.md
AGENTS.md instructions for yfedoseev/browser_oxide, covering browseroxide — conventions for ai coding assistants, build & test, architecture, key conventions and where to read more.