Charlotte
Turn arbitrary web pages into structured, agent-readable content using headless Chromium inside MCP workflows.
Overview
Charlotte is a MCP server for the Build phase that renders web pages with headless Chromium into structured, agent-readable representations.
What is this MCP server?
- npm package @ticktockbent/charlotte (version 0.1.3) with stdio MCP transport
- Headless Chromium renders pages into structured representations
- Output tuned for agent consumption rather than raw HTML dumps
- stdio server—runs locally next to Claude Code or Cursor
- Source on GitHub: ticktockbent/charlotte
- npm package version 0.1.3 (@ticktockbent/charlotte)
- Transport: stdio
- Rendering engine: headless Chromium (per server description)
Community signal: 147 GitHub stars.
What problem does it solve?
Agents choking on dynamic sites or messy HTML slows research, QA, and integration tasks that need what users actually see in the browser.
Who is it for?
Solo builders who need dependable headless browsing inside Claude Code or Cursor for docs, competitors, or JS-heavy app surfaces.
Skip if: Large-scale unattended scraping pipelines or teams that cannot run Chromium locally alongside the agent.
What do I get? / Deliverables
After npm install and stdio MCP registration, your agent can request rendered page structure suitable for analysis and follow-on build steps in the same session.
- Rendered web page content in structured agent-readable form
- MCP stdio bridge replacing ad-hoc Playwright scripts for one-off page reads
- Repeatable page fetch workflow inside agent sessions
Recommended MCP Servers
Journey fit
Fetching and normalizing live web UI happens while building features that depend on page structure—scrapers, monitors, and agent research loops—not only at launch SEO tuning. Integrations subphase fits browser rendering bridges that connect external sites to your agent’s context window without custom Puppeteer scripts each time.
How it compares
Headless browser rendering MCP, not a SEO rank tracker or passive markdown fetch skill.
Common Questions / FAQ
Who is Charlotte MCP for?
Indie developers and agent builders who want headless Chromium page renders exposed as MCP tools instead of one-off browser automation repos.
When should I use Charlotte MCP?
Use it during build integrations when you need structured page content for competitor review, UI QA, or feeding context before you implement features.
How do I add Charlotte to my agent?
Install @ticktockbent/charlotte from npm (0.1.3) and add the stdio MCP server configuration in Claude Code, Cursor, or another MCP-capable client per package docs.