CatFee Dokobot

Use dokotom (browser automation via real Chrome) to read web pages, extract JS-rendered content, and monitor dynamic data. Use when you need to fetch content...

Security Scan

The instructions match the stated purpose (using a real Chrome via Dokobot) but the skill metadata omits the real prerequisites and it instructs installing a third‑party CLI + browser extension that can access your authenticated browser sessions — review and limit exposure before installing.

Purpose & Capability

The SKILL.md clearly requires the @dokobot/cli, a Dokobot Chrome extension, and a local bridge, but the registry metadata lists no required binaries/config — this is an incoherence: the skill claims no prerequisites while runtime instructions depend on external software and Chrome.

Instruction Scope

Runtime instructions are narrowly scoped to using dokobot to read pages via a local Chrome bridge and parsing its output. They do not instruct reading unrelated host files or env vars. However the instructions explicitly rely on the user's Chrome session (cookies/auth) for pages that require login — this elevates sensitivity of the data accessed.

Install Mechanism

There is no formal install spec in the registry; instead SKILL.md tells users to run `npm install -g @dokobot/cli`, install a Chrome extension, and run `dokobot install-bridge`. Installing a third-party npm package and browser extension is a moderate risk and should be verified (package source, publisher, extension permissions).

Credentials

No environment variables are requested, which is consistent, but the skill expects to access the user's live Chrome session (including cookies and logged‑in state) to read pages requiring login. That implicit access to private/authenticated data is a significant sensitivity that is not reflected in the metadata or warned about.

Persistence & Privilege

The skill is not always-enabled and does not request system-wide privileges in metadata. The only persistence-related action suggested is running a local bridge (`dokobot install-bridge`), which may create a local service; this is expected for the stated functionality but should be reviewed by the user.

What to consider before installing

This skill is coherent with its goal (driving a real Chrome to render JS pages) but exercise caution. Before installing: (1) verify the npm package and Dokobot extension source (check npm and the extension store or project GitHub) and confirm you trust the publisher; (2) avoid using your primary browser profile — create a fresh profile or a disposable browser/VM so the extension/bridge cannot access your regular cookies and logins; (3) review extension permissions and the npm package's repository and recent release history; (4) if you must access pages behind login, prefer providing a dedicated test account rather than your personal accounts; (5) be cautious about running `dokobot install-bridge` on production systems. These steps reduce the risk that a third‑party CLI/extension will exfiltrate sensitive browsing data.

CatFee Dokobot Skill

Use dokobot to read web pages through a real Chrome browser, enabling JavaScript rendering for dynamic/SPA pages.

Prerequisites

• dokobot CLI installed globally: npm install -g @dokobot/cli
• Chrome browser with Dokobot extension installed
• Local bridge running: dokobot install-bridge (one-time setup)
• Chrome must be open with the Dokobot extension enabled for --local mode to work

Core Commands

Check Bridge Status

dokobot doko list

Shows connected browsers. If empty, open Chrome with the Dokobot extension.

Read a Web Page (Local Mode - Free)

dokobot read --local "<URL>" --timeout 60000

Reads a page through local Chrome. Wait for --timeout (default 30s) to let JS render.

For dynamic pages (JS-rendered data):

dokobot read --local "<URL>" --timeout 60000

Use --timeout 60000 (60s) for pages with heavy JS loading. For very slow pages, increase to 90000.

To continue a session (for paginated content):

dokobot read "<URL>" --session-id <SESSION_ID> --screens 5

Close a Session

dokobot doko close <SESSION_ID>

Workflow

1. Check bridge is running: dokobot doko list
2. If no devices: Open Chrome with Dokobot extension, wait a few seconds, retry
3. Read page: dokobot read --local "<URL>" --timeout 60000
4. Parse output: Extract the data needed from the markdown output
5. Present results: Format clearly for the user

Tips

• Delay for dynamic content: Use --timeout 60000 or higher. Pages like East Money (东方财富) data centers need 45-60 seconds to render.
• Chrome not responding: If bridge shows device but read fails, close and reopen Chrome
• Session continuity: For multi-page reading, use --session-id from previous read to continue
• Headless limitation: If Chrome is closed or the extension is disabled, local mode won't work

Common Sites

Site	URL Pattern	Notes
东方财富 行情中心	https://quote.eastmoney.com/changes	盘口异动，60s timeout
东方财富 龙虎榜	https://data.eastmoney.com/stock/lhb.html	60s timeout
东方财富 个股资金流向	https://data.eastmoney.com/zjlx/<CODE>.html	CODE = stock code without market prefix
东方财富 数据中心	https://data.eastmoney.com/	60s timeout

Troubleshooting

Problem	Solution
"No available devices"	Open Chrome with Dokobot extension enabled
"Bridge not running"	Run dokobot install-bridge
Data missing/incomplete	Increase --timeout to 90000
Page shows "not found"	URL may have changed; search for correct URL
Read times out	Use longer --timeout or try again

Example Workflows

Monitor Stock Changes (Every 30s × 6)

for ($i = 1; $i -le 6; $i++) {
    Write-Host "=== 第${i}次刷新 $(Get-Date -Format 'HH:mm:ss') ==="
    $result = dokobot read --local "<URL>" --timeout 60000 2>&1
    # Parse relevant lines
    if ($i -lt 6) { Start-Sleep -Seconds 30 }
}

Read Dynamic Data Page

dokobot read --local "https://data.eastmoney.com/stock/lhb.html" --timeout 60000