القائمة
Browse the web using Plasmate for 16x fewer tokens than raw HTML. Use when fetching documentation, reading web pages, extracting structured data, or interacting with websites.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
| name | plasmate |
| description | Browse the web using Plasmate for 16x fewer tokens than raw HTML. Use when fetching documentation, reading web pages, extracting structured data, or interacting with websites. |
| config | {"mcp":{"command":"npx","args":["-y","plasmate-mcp"]}} |
Browse web pages using Plasmate's Semantic Object Model (SOM) instead of raw HTML. SOM compresses web content by 16.6x on average while preserving page structure, element roles, and interactive affordances.
Once the MCP server is connected, you have these tools:
Fetch any URL and get the Semantic Object Model (SOM). Returns structured JSON with semantic regions (navigation, main, footer), typed elements (links, buttons, headings, tables), and available actions.
fetch_page(url: "https://example.com")
Get clean, readable text from any URL. Use this when you only need the text content without structural metadata.
extract_text(url: "https://docs.example.com/api")
Open a persistent browser session for multi-step interactions.
open_page(url: "https://example.com/login")
Click an element by its SOM ID within an open page session.
click(session_id: "...", element_id: "btn-submit")
Execute JavaScript in an open page session.
evaluate(session_id: "...", expression: "document.title")
Close a browser session when done.
close_page(session_id: "...")
The skill auto-configures the MCP server. If you need to install Plasmate manually:
# Install the MCP wrapper
npm install -g plasmate-mcp
# Install the Plasmate engine (pick one)
brew tap plasmate-labs/plasmate && brew install plasmate
# or: cargo install plasmate
# or: pip install plasmate
Then add the MCP server:
claude mcp add plasmate -- npx plasmate-mcp
extract_text for simple reading tasks; it returns fewer tokens than fetch_pagefetch_page when you need to understand page structure or find interactive elementsopen_page + click for multi-step workflows like form filling or navigationPLASMATE_CACHE_API_KEY environment variable for faster cached lookups via the SOM Cache CDN