MeshArc — Claude Code MCP server: what it does, how to install
MeshArc lets an agent collect website pages as clean Markdown and track changes across crawls. Use it to inspect a site or monitor how its pages evolve.
claude mcp add --transport http mesharc https://mcp.mesharc.dev/mcpWhat MeshArc does
MeshArc gives an agent tools to search for pages, read them, and crawl a website. Page content can be returned as Markdown, and the service records pages that were added, removed, or changed between crawls.
A one-off crawl can be kept as a project with a schedule for continued monitoring. One-off crawls expire after 30 days, so ongoing tracking requires keeping the crawl as a project. Claude Code connects to the MCP server with a single command.
What it helps with
- Extract page content as Markdown
- Crawl a website to collect its pages
- Find and read pages matching a search
- Track pages added, removed, or changed
Example prompts for the agent
- Collect the pages on this site as Markdown.
- Find pages about Python packaging and read the results.
- Track changes to this site across crawls.
How to install MeshArc
Platform CLI: remote server
claude mcp add --transport http mesharc https://mcp.mesharc.dev/mcpPlatform CLI: remote server
codex mcp add mesharc --url https://mcp.mesharc.dev/mcpConfig file
{
"mcpServers": {
"mesharc": {
"url": "https://mcp.mesharc.dev/mcp"
}
}
}Config file
{
"servers": {
"mesharc": {
"type": "http",
"url": "https://mcp.mesharc.dev/mcp"
}
}
}Platform CLI: remote server
gemini mcp add -t http mesharc https://mcp.mesharc.dev/mcpStep by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about MeshArc
How do I install MeshArc in Claude Code?
Run claude mcp add --transport http mesharc https://mcp.mesharc.dev/mcp.
What does MeshArc do?
It lets an agent search for and read web pages, crawl sites, and retrieve page content as Markdown. It can also record page changes between crawls.
How does MeshArc monitor a website over time?
Keep a crawl as a project and set a schedule. One-off crawls expire after 30 days.
