mcp · siteglassio — Claude Code MCP server: site checks and setup
Scan a web app, archive public pages, or run E2E flows against a deployed site. Give your agent tools to check changes in the browser.
claude mcp add --env SITEGLASS_API_KEY=<SITEGLASS_API_KEY> --env SITEGLASS_BASE_URL=<SITEGLASS_BASE_URL> --transport stdio siteglassio-mcp -- npx -y siteglass-mcpWhat mcp does
siteglass-mcp provides tools for inspecting web apps and capturing pages. siteglass_scan_local opens a development app in a browser on your machine and reports broken links or assets, console errors, accessibility findings, and SEO issues. For a public page, siteglass_render can create an rrweb capture, PDF, or PNG.
For a deployed site, the server can register and verify ownership, scan pages, derive E2E flows, and run them. Flow results include step status, screenshots, video, and an rrweb replay; test credentials can be supplied for authenticated flows. Longer operations return an ID that the agent can check with a matching get_* tool.
Local scans need no authentication. Hosted tools create and cache an API key at ~/.siteglass/key, or can use SITEGLASS_API_KEY. The first scan for a site is free; later scans and flow runs use credits from a prepaid balance. A site that blocks datacenter IPs can return an empty capture.
What it helps with
- find broken links and console errors in a development app
- check pages for accessibility and SEO issues
- save a public page as a PDF, PNG, or rrweb capture
- derive E2E flows from a site scan
- review flow steps with screenshots and video
Example prompts for the agent
- Scan my app and report the issues you find.
- Archive this public page as a PDF.
- Generate E2E flows for the site and run them with test credentials.
How to install mcp
Platform CLI
claude mcp add --env SITEGLASS_API_KEY=<SITEGLASS_API_KEY> --env SITEGLASS_BASE_URL=<SITEGLASS_BASE_URL> --transport stdio siteglassio-mcp -- npx -y siteglass-mcpReplace <…> with your own values
Platform CLI
codex mcp add siteglassio-mcp --env SITEGLASS_API_KEY=<SITEGLASS_API_KEY> --env SITEGLASS_BASE_URL=<SITEGLASS_BASE_URL> -- npx -y siteglass-mcpReplace <…> with your own values
Config file
{
"mcpServers": {
"siteglassio-mcp": {
"command": "npx",
"args": [
"-y",
"siteglass-mcp"
],
"env": {
"SITEGLASS_API_KEY": "<SITEGLASS_API_KEY>",
"SITEGLASS_BASE_URL": "<SITEGLASS_BASE_URL>"
}
}
}
}Replace <…> with your own values
Config file
{
"servers": {
"siteglassio-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"siteglass-mcp"
],
"env": {
"SITEGLASS_API_KEY": "<SITEGLASS_API_KEY>",
"SITEGLASS_BASE_URL": "<SITEGLASS_BASE_URL>"
}
}
}
}Replace <…> with your own values
Platform CLI
gemini mcp add -e SITEGLASS_API_KEY=<SITEGLASS_API_KEY> -e SITEGLASS_BASE_URL=<SITEGLASS_BASE_URL> siteglassio-mcp npx -y siteglass-mcpReplace <…> with your own values
Step 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 mcp
How do I install mcp in Claude Code?
Run claude mcp add mcp -e SITEGLASS_API_KEY=<SITEGLASS_API_KEY> -- npx -y siteglass-mcp. Provide your API key.
What does mcp do?
It scans web apps, captures public pages, and runs E2E flows against deployed sites. Local scans use a browser running on your machine.
Does a local scan require an API key?
No. siteglass_scan_local runs without authentication. Hosted tools use an API key, which you can provide through SITEGLASS_API_KEY.
Similar MCP servers
All MCP servers for Testing and QA →mcp by other authors comes first.
