Fetch Rendered Page

Fetches fully rendered HTML from JavaScript-heavy pages.

When to Use

• Standard WebFetch returns empty/incomplete content
• Page is a Single Page Application (SPA)
• Content renders client-side (React, Vue, Angular)
• Need text from modern documentation sites

Setup

Install dependencies once:

cd ${CLAUDE_SKILL_DIR}
npm install
npx playwright install chromium

Usage

Start Server

cd ${CLAUDE_SKILL_DIR} && node server.js &
sleep 3

Fetch Page

curl -s "http://localhost:3456/render?url=$ARGUMENTS&waitTime=5000"

Extract Text

curl -s "http://localhost:3456/render?url=$ARGUMENTS" | jq -r '.textContent'

Stop Server

pkill -f "node.*server.js"

Parameters

Param	Description	Default
url	Target URL	required
waitTime	Wait after load (ms)	3000
waitFor	CSS selector to wait for	-
timeout	Page timeout (ms)	30000

Example

Input: /fetch-rendered https://open.larksuite.com/document/client-docs/bot-v3/add-custom-bot

cd ${CLAUDE_SKILL_DIR} && node server.js &
sleep 3
curl -s "http://localhost:3456/render?url=https://open.larksuite.com/document/client-docs/bot-v3/add-custom-bot&waitTime=5000" | jq -r '.textContent'
pkill -f "node.*server.js"

Errors

• Server won't start: Run npm install in skill directory
• Empty content: Increase waitTime to 5000-10000
• Port in use: lsof -ti:3456 | xargs kill -9