browser_cdp
当用户明确希望连接到已运行的 Chrome 浏览器(connect_cdp)、扫描本地 CDP 端口、或以暴露 CDP 端口的方式启动浏览器(start + cdp_port)时,使用本 skill。用户没有明确提到 CDP、共享浏览器或远程调试时,启动浏览器不得携带 cdp_port,也不得使用 connect_cdp。CDP 模式会暴露浏览器历史、Cookies 等敏感信息,使用前须告知用户;同一 workspace 同时只能运行或连接一个浏览器。
Supply asset profile
Coding and developer agents
Code review, repo analysis, testing, CI, GitHub, DevOps, and developer workflow skills.
Scenario
Testing and QA
I need my agent to test a web app, reproduce bugs, and verify fixes.
Agent fit
Claude Code + Browser agents + CLI
Codex, Claude Code, Cursor, CLI, or custom agents.
Install
Ready
npx skills add hyqibot/token-free-openclaw --skill browser_cdp
Maintenance
fresh
1d since push
Risk
Needs review
Dependency or permission surface needs review
GitHub quality
116
67/100 Quality · 70/100 Trust
Coverage tags
Review notes
Dependency or permission surface needs review · Permission surface may require sandboxing
Agent adoption scorecard
Trust, audit, and install readiness at a glance
These scores combine public repository metadata, OpenAgentSkill review signals, maintenance freshness, and install readiness. They are a shortlist signal, not a replacement for human review.
Quality
PromisingUseful candidate, but compare it with alternatives before adopting.
Trust
Sandbox onlyUseful candidate with missing or mixed trust signals. Keep it in an isolated workspace until the outcome loop proves task fit.
Audit
Needs reviewA machine-readable review of install readiness, security metadata, maintenance, and adoption risk.
OpenAgentSkill Trust Score v5
Human review before install
Run only in a sandbox and compare close alternatives before using it for real work.
Stars
116 GitHub stars
Repo activity
116 stars, 32 forks
Maintenance
1d since push
License
Apache-2.0
Install
npx skills add hyqibot/token-free-openclaw --skill browser_cdp
Install safety
credential-bearing install command, standard package or runtime install path
Permission surface
secrets or environment access, shell or command execution
Agent outcomes
No agent outcome data yet
Docs
Usable metadata, review docs
Risk summary
Review before production
- Quality score needs review
- Permission surface needs review: secrets or environment access, shell or command execution
- Stars/forks activity: 116 stars, 32 forks; issue activity unavailable in current metadata
- Dependency/runtime risk: command execution surface, credential or environment access
Install readiness
Install path available
- Install path is available
- Repository evidence is available
- License is declared
- No Agent Proven outcome evidence yet
Agent-readable metadata
Machine-readable decision data for this skill.
Use this block or the embedded JSON to decide whether an agent should install this skill, choose an alternative, or ask for human review first.
Suited tasks
- Browser automation workflows
- Claude Code teams
- builders willing to evaluate younger projects
- Navigate pages
Suited agents
Install decision
- Command
- npx skills add hyqibot/token-free-openclaw --skill browser_cdp
- Policy
- block
- Human review
- yes
Trust and risk
- Trust
- 62/100
- Audit
- 76/100
- Risk level
- Needs review
Outcome loop
- Endpoint
- /api/agent/outcome
- Event ID
- resolve
- Outcomes
- 5
Install command
npx skills add hyqibot/token-free-openclaw --skill browser_cdpDo not use when
- teams that need a vendor-supported SLA
- high-compliance environments without internal security review
- No major risk signals from current metadata
- High-risk permission hints: Shell or command execution, Secrets or environment access
- Dependency or permission surface needs review
Agent safety v2
32/100 · Avoid automatic install
This skill should not be selected by an agent without explicit human security review.
Do not auto-install. Inspect the source, dependencies, and permission surface first.
high
Shell or command execution
Skill metadata references terminal, CLI, shell, subprocess, or command execution workflows.
medium
Browser automation
Skill may drive a browser or interact with web pages.
medium
Network access
Skill likely fetches remote pages, APIs, repositories, or external services.
medium
Filesystem access
Skill may read or write project files, documents, generated artifacts, or local workspace state.
- High-risk permission hints: Shell or command execution, Secrets or environment access
- Dependency or permission surface needs review
Install targets
Install this skill in your agent workflow
Use the public install endpoint to fetch the command, safety checklist, target prompts, and canonical links for this skill.
OpenAgentSkill CLI
Resolve policy, run the source installer safely, and report a verified install receipt.
$ npx --yes https://github.com/Leon-Drq/openagentskill/releases/download/cli-v0.2.1/openagentskill-0.2.1.tgz install hyqibot-browser-cdpAgent resolve plan
Let an agent verify fit before installing.
The Resolve API returns the selected skill, alternatives, safety policy, audit notes, install target, and copy-paste prompt an agent can follow without scraping this page.
Open JSON
/api/agent/resolve?task=Use%20browser_cdp%20for%20an%20agent%20workflow&agent=codex&max_risk=medium
Resolve text
/api/agent/resolve?task=Use%20browser_cdp%20for%20an%20agent%20workflow&agent=codex&max_risk=medium&format=text
Install handoff
/api/skills/hyqibot-browser-cdp/install
Agent should check
- Task fit and alternatives from Resolve API.
- Audit score, trust score, and safety policy warnings.
- Install target compatibility for Codex, Claude Code, Cursor, or CLI.
Copy prompt
Task: Use browser_cdp in this workspace.
Resolve first: https://www.openagentskill.com/api/agent/resolve?task=Use%20browser_cdp%20for%20an%20agent%20workflow&agent=codex&max_risk=medium
Review install handoff: https://www.openagentskill.com/api/skills/hyqibot-browser-cdp/install
Install command: npx skills add hyqibot/token-free-openclaw --skill browser_cdp
Before running it, summarize audit warnings, required permissions, and the fallback skill if install is risky.Agent handoff
Give an agent the install path, not another directory page.
Use the public install endpoint to fetch the command, safety checklist, target prompts, and canonical links for this skill.
Install handoff
/api/skills/hyqibot-browser-cdp/install
LLM text format
/api/skills/hyqibot-browser-cdp/install?format=text
Find alternatives
/api/skills/search?q=browser_cdp&limit=3
Agent prompt
Use browser_cdp for this task. Review https://www.openagentskill.com/api/skills/hyqibot-browser-cdp/install, then install with: npx skills add hyqibot/token-free-openclaw --skill browser_cdpRegistry metadata
Agent-readable profile for automatic skill selection.
This page exposes the same decision, trust, audit, use-case, and install signals through the Registry API, so agents can rank this skill without scraping the UI.
Manifest
/api/registry/manifest/hyqibot-browser-cdp
LLM text
/api/registry/manifest/hyqibot-browser-cdp?format=text
Install alias
/api/registry/install/hyqibot-browser-cdp
Recommend
/api/registry/recommend?task=Use%20browser_cdp%20in%20an%20agent%20workflow&limit=3
Agent fit
Browser automation
Use-case tags
Platforms
Claude Code, Browser agents
Audit report
Needs review · 76/100
A machine-readable review of install readiness, security metadata, maintenance, and adoption risk.
Agent decision cockpit
Fallback candidate for Browser automation
Prototype with this skill first; keep a fallback candidate ready.
Role in stack
Fallback candidate
Primary fit
Browser automation
Trust label
Prototype first
Install path
Command ready
Use when
- Browser automation workflows
- Claude Code teams
- builders willing to evaluate younger projects
Evidence
- recent repository activity
- install command or GitHub repo available
- 67/100 quality profile
- 2 OpenAgentSkill engagement events
review first
- No major risk signals from current metadata
Implementation path
- 1Install it in a sandbox agent and run one Browser automation task end to end.
- 2Compare output quality, latency, and failure behavior against at least one alternative.
- 3Promote it into production only after reviewing repository permissions, license, and maintenance signals.
Trust profile
Sandbox only
Useful candidate with missing or mixed trust signals. Keep it in an isolated workspace until the outcome loop proves task fit.
GitHub adoption
INFO116 GitHub stars
Stars/forks activity
CHECK116 stars, 32 forks; issue activity unavailable in current metadata
Recent maintenance
PASS1d since push
License clarity
PASSApache-2.0
Good signals
- AI review approved
- Install path is available
- Repository evidence is available
- Recently maintained repository
- Install command has no obvious high-risk pattern
- Outcome loop is ready but needs first real agent run
Review before install
- Quality score needs review
- Permission surface needs review: secrets or environment access, shell or command execution
- Stars/forks activity: 116 stars, 32 forks; issue activity unavailable in current metadata
- Dependency/runtime risk: command execution surface, credential or environment access
- Permission surface: secrets or environment access, shell or command execution
- No real agent outcome reports yet
- Human review required before unattended installation
Recommended action
Run only in a sandbox and compare close alternatives before using it for real work.
Quality profile
Promising candidate for agent workflows
Useful candidate, but compare it with alternatives before adopting.
Workflow fit
Use this skill in these scenarios
Operate web apps
Browser automation
I need my agent to control a browser, fill forms, and verify web app workflows.
Verify behavior
Testing and QA
I need my agent to test a web app, reproduce bugs, and verify fixes.
Operate local tools
Local desktop
I need my agent to operate local files and desktop apps in a repeatable workflow.
Workflow fit
Add it to a complete workflow
Operate and verify web apps
Browser QA agent
A workflow for agents that navigate products, fill forms, take screenshots, and verify real user flows across web applications.
Scrape, clean, and reuse web data
Web data pipeline
A practical workflow for agents that crawl public pages, extract clean content, normalize data, and hand it to downstream research or RAG workflows.
Turn skills into distribution
Content growth agent
A workflow for turning newly indexed skills into SEO briefs, social drafts, comparison pages, and reusable publishing workflows.
Alternative shortlist
Compare before you install
Similar skills that may fit this task.
UI-TARS Desktop
Run multimodal agents that operate desktop interfaces
MoneyPrinterTurbo
利用AI大模型,一键生成高清短视频 Generate short videos with one click using AI LLM.
Cua
Open-source infrastructure for Computer-Use Agents. Sandboxes, SDKs, and benchmarks to train and evaluate AI agents that can control full desktops (macOS, Linux, Windows).
Overview
--- name: browser_cdp description: "当用户明确希望连接到已运行的 Chrome 浏览器(connect_cdp)、扫描本地 CDP 端口、或以暴露 CDP 端口的方式启动浏览器(start + cdp_port)时,使用本 skill。用户没有明确提到 CDP、共享浏览器或远程调试时,启动浏览器不得携带 cdp_port,也不得使用 connect_cdp。CDP 模式会暴露浏览器历史、Cookies 等敏感信息,使用前须告知用户;同一 workspace 同时只能运行或连接一个浏览器。" metadata: builtin_skill_version: "1.1" copaw: emoji: "🔌" requires: {} ---
# 浏览器 CDP 连接参考
**Chrome DevTools Protocol(CDP)** 允许外部程序连接并控制一个已在运行的 Chrome 进程。本 skill 覆盖三类场景:
1. **扫描本地 CDP 端口** — 发现正在运行的可连接 Chrome 2. **连接已有 Chrome** — 附加到外部 Chrome,不影响其运行 3. **启动暴露 CDP 端口的浏览器** — 让 Playwright 启动的浏览器对外可见
> **⚠️ 隐私提示:使用 CDP 前请知晓** > > - **默认模式**(`start` 不带 `cdp_port`):浏览器完全由 Playwright 私有管理,历史记录、Cookies、登录态**不会暴露**给任何外部程序。 > - **CDP 模式**(`start` + `cdp_port` 或 `connect_cdp`):任何能访问该端口的程序都可以读取浏览器的**完整历史记录、Cookies、当前页面内容、已保存密码**等敏感信息。仅在信任的本地环境中使用,不要在公共网络或多用户服务器上暴露 CDP 端口。
> **⚠️ 单实例限制:一个 workspace 同时只能运行/连接一个浏览器** > > 无论是 Playwright 启动的浏览器还是 CDP 连接,同一 workspace 内同一时间只允许存在一个。需要切换时,必须先执行 `stop`,再启动或连接新的浏览器。
---
## 何时使用
**仅在用户明确表达以下意图时才使用本 skill:**
- 用户说:「连接到我已打开的 Chrome」「扫描一下本地有没有可连接的浏览器」 - 用户明确希望多个 agent 或外部工具**共享**同一个浏览器实例 - 用户明确要求浏览器**对外可见/可调试** - 用户主动提到 CDP、远程调试端口等概念
**以下情况不要使用 CDP 模式,直接用普通 `start` 即可:**
- 用户只是说「打开浏览器」「帮我打开某网站」,没有提到共享或调试 - 用户没有明确说明需要暴露浏览器给其他程序 - 不确定用户是否了解 CDP 的隐私风险
**使用前须告知用户 CDP 模式会暴露浏览器历史、Cookies 等敏感信息,确认用户知情后再操作。**
---
## 场景一:扫描本地 CDP 端口
默认扫描端口范围 **9000–10000**,并发探测,速度很快。
```json {"action": "list_cdp_targets"} ```
指定单个端口: ```json {"action": "list_cdp_targets", "port": 9222} ```
自定义扫描范围(未找到时可扩大范围): ```json {"action": "list_cdp_targets", "port_min": 8000, "port_max": 12000} ```
**成功返回示例:** ```json { "ok": true, "found": { "9222": [{"title": "New Tab", "url": "chrome://newtab/", ...}] }, "message": "Found CDP endpoints on port(s): 9222" } ```
**未找到时** 返回 `ok: false`,并提示扩大范围或确认 Chrome 是否以 `--remote-debugging-port=N` 启动。
---
## 场景二:连接已有 Chrome(connect_cdp)
先扫描到端口后,再连接:
```json {"action": "connect_cdp", "cdp_url": "http://localhost:9222"} ```
- 连接成功后可正常使用 `open`、`snapshot`、`click`、`type` 等所有操作 - **不会影响 Chrome 进程**:执行 `stop` 时只断开 Playwright 连接,Chrome 继续运行 - 连接期间每次操作前会自动检查连接是否正常,断开时返回错误提示重新连接 - **隐私风险**:连接后 agent 可读取该 Chrome 的完整 Cookies、历史记录、当前页面内容,请确认用户知情 - **单实例**:如果当前已有浏览器在运行(无论是 Playwright 启动的还是另一个 CDP 连接),必须先 `stop` 再连接
---
## 场景三:启动带 CDP 端口的浏览器
让 Playwright 启动浏览器时暴露指定 CDP 端口,其他工具可同时连接:
```json {"action": "start", "cdp_port": 9222} ```
启动成功后返回: ```json { "ok": true, "message": "Browser started with CDP port 9222", "cdp_url": "http://localhost:9222" } ```
之后可用 `list_cdp_targets` 验证端口已暴露,或将 `cdp_url` 提供给其他 agent / 工具连接。
**注意:** - **隐私风险**:暴露端口后,任何能访问该端口的程序均可读取浏览器历史、Cookies 及页面内容,操作前请告知用户 - **单实例**:当前 workspace 已有浏览器运行时,无法再启动新浏览器,必须先 `stop`
---
## Cookies 与数据持久化
三种启动方式都复用同一个 workspace 的 `user_data_dir`:
| 启动方式 | Cookies 复用 | 对外可访问 | |---|---|---| | `start`(默认) | ✅ | ❌ | | `start` + `cdp_port` | ✅ | ✅ | | `connect_cdp`(已有 Chrome) | ✅(若该 Chrome 使用了相同目录) | ✅ |
---
## stop 行为说明
两种 CDP 模式对 `stop` 的响应截然不同:
- **`connect_cdp`**:agent 附加到用户已有的 Chrome,**不拥有**该进程。执行 `stop` 只断开 Playwright 连接,Chrome 继续运行,用户页面不受影响。 - **`start` + `cdp_port`**:agent 自行启动并管理浏览器。执行 `stop` 会**终止 Chrome 进程**,其他通过该 CDP 端口连接的外部工具也会断线。
| 当前状态 | stop 效果 | |---|---| | CDP 连接(`connect_cdp`) | 仅断开 Playwright 连接,**Chrome 进程继续运行** | | Playwright 启动的浏览器(`start` + `cdp_port`) | **终止 Chrome 进程**,外部 CDP 连接同时断线 |
---
## 清除缓存
```json {"action": "clear_browser_cache"} ```
- 浏览器运行中:通过 CDP 清除 HTTP 缓存,**无需重启** - 浏览器已停止:删除磁盘上的缓存目录 - Cookies 和 Local Storage **不受影响**
---
## 常见问题
**Chrome 启动后无法扫描到 CDP 端口?**
Chrome 必须以独立进程 + 指定 `user-data-dir` 启动,否则新实例会被移交给已有进程:
```bash pkill -x "Google Chrome" # 先关闭已有 Chrome /Applications/Google\ Chrome.app/Contents/MacOS/Google\ Chrome \ --remote-debugging-port=9222 \ --user-data-dir=/tmp/chrome-debug-9222 \ --no-first-run ```
**CDP 连接中断后怎么恢复?**
任何操作都会返回: ```json {"ok": false, "error": "CDP connection lost (was: http://localhost:9222). Reconnect with action='connect_cdp'."} ``` 按提示重新执行 `connect_cdp` 即可。
Technical details
- Version
- 1.0.0
- License
- Apache-2.0
- Last updated
- Aug 21, 2026
- Published
- Aug 21, 2026
Decision snapshot
Fallback candidate
recent repository activity
Audit
Install review
Install and adoption review
- Security
- 72/100
- Maintenance
- 100/100
- Install
- 92/100
Agent-proven evidence
Agent-proven evidence
Outcome reports after resolve, review, install, and one narrow run.
- Success rate
- —
- Recent failure
- —
- Outcomes
- 0
- Output quality
- —
- Failed
- 0
- Not relevant
- 0
- Installs
- 0
- Risk blocked
- 0
- Setup needed
- 0
- Production
- 0
No agent outcome data yet. The first agent run can report success, setup needs, risk blocks, failure, or not-relevant through /api/agent/outcome.
Install
Add to agent workflow
Free and open source. Review the report before installing into production agents.
Growth loop
Share kit
Scenario-led draft for browser_cdp, ready for a manual X post.
browser_cdp: 当用户明确希望连接到已运行的 Chrome 浏览器(connect_cdp)、扫描本地 CDP 端口、或以暴露 CDP 端口的方式启动浏览器(start + cdp_port)时,使用本... 116 stars https://www.openagentskill.com/skills/hyqibot-browser-cdp?ref=x
Optional reply with install command
Listing + install path for browser_cdp: https://www.openagentskill.com/skills/hyqibot-browser-cdp?ref=x Install: npx skills add hyqibot/token-free-openclaw --skill browser_cdp
Listing source
Registry indexed
This listing was indexed from public sources and is not marked official until a maintainer claim is approved.
- Creator
- hyqibot
- Indexed by
- OpenAgentSkill community index
Attribution links to the public repository or creator profile. Creators can claim the listing to update ownership signals.
Claim this skillOwner claim
Claim this skill listing
This Registry indexed listing is attributed to hyqibot but is not marked official yet. Claim it to add a verified owner signal and make future launch, install, and audit updates easier to trust.
Creator backlink kit
Add the evidence badges to your README
Show the canonical listing, current trust and audit signals, and real Agent-Proven evidence where developers evaluate the repository.
[](https://www.openagentskill.com/skills/hyqibot-browser-cdp)
[](https://www.openagentskill.com/skills/hyqibot-browser-cdp)
[](https://www.openagentskill.com/skills/hyqibot-browser-cdp/audit)
[](https://www.openagentskill.com/skills/hyqibot-browser-cdp)Author
hyqibot
@hyqibot
Tags
Platform fit
Health signals
- GitHub stars
- 116
- Quality score
- 38/100
- Last GitHub push
- Aug 21, 2026
- Framework hints
- Unknown
- OpenAgentSkill views
- 2
- Install copies
- 0
- Outbound clicks
- 0
Community signal
Share whether this skill looks useful for your agent workflow. Aggregated feedback improves rankings over time.
Trust & safety
Sandbox only
- GitHub adoption116 GitHub starsINFO
- Stars/forks activity116 stars, 32 forks; issue activity unavailable in current metadataCHECK
- Recent maintenance1d since pushPASS
- License clarityApache-2.0PASS
- README/SKILL.md completenessPublic metadata needs stronger README/SKILL.md contextINFO
- Dependency/runtime riskcommand execution surface, credential or environment accessCHECK
Related skills
UI-TARS Desktop
Run multimodal agents that operate desktop interfaces
37.0K StarsMoneyPrinterTurbo
利用AI大模型,一键生成高清短视频 Generate short videos with one click using AI LLM.
88.5K StarsCua
Open-source infrastructure for Computer-Use Agents. Sandboxes, SDKs, and benchmarks to train and evaluate AI agents that can control full desktops (macOS, Linux, Windows).
21.4K Stars