Squally — Claude Code MCP server for Playwright analytics
This MCP server for Claude Code reads test analytics from Squally. It surfaces flaky tests, failure causes, and CI time spent on retries.
MIT updated 2026-11-16
claude mcp add --env SQUALLY_API_KEY=<SQUALLY_API_KEY> --env SQUALLY_API_URL=<SQUALLY_API_URL> --env SQUALLY_USE_SYSTEM_CA=<SQUALLY_USE_SYSTEM_CA> --transport stdio squally -- npx -y squally-mcpWhat Squally does
Squally gives an agent access to Playwright run data and per-test results. It can locate a run, inspect a failing test’s error and stack, check test stability, and group recurring error signatures across runs.
The server runs locally over stdio and connects to Squally’s API over HTTPS. Access is read-only: an organization key covers that organization’s projects, but does not expose secrets, webhook URLs, screenshots, videos, or traces. It requires Node 22 or newer, an organization read key, and Squally’s Standard plan. A project ingest key used to send Playwright results is rejected.
What it helps with
- find tests that pass only after a retry
- inspect why a test failed in a particular run
- measure CI time lost to retries and failures
- group recurring errors across test runs
Example prompts for the agent
- Find flaky tests in the project and show how often they retry.
- Investigate the failure of
checkout.spec.tsin the latest run on the main branch. - Group errors from the period and show which tests fail most often.
How to install Squally
Platform CLI
claude mcp add --env SQUALLY_API_KEY=<SQUALLY_API_KEY> --env SQUALLY_API_URL=<SQUALLY_API_URL> --env SQUALLY_USE_SYSTEM_CA=<SQUALLY_USE_SYSTEM_CA> --transport stdio squally -- npx -y squally-mcpReplace <…> with your own values
Platform CLI
codex mcp add squally --env SQUALLY_API_KEY=<SQUALLY_API_KEY> --env SQUALLY_API_URL=<SQUALLY_API_URL> --env SQUALLY_USE_SYSTEM_CA=<SQUALLY_USE_SYSTEM_CA> -- npx -y squally-mcpReplace <…> with your own values
Config file
{
"mcpServers": {
"squally": {
"command": "npx",
"args": [
"-y",
"squally-mcp"
],
"env": {
"SQUALLY_API_KEY": "<SQUALLY_API_KEY>",
"SQUALLY_API_URL": "<SQUALLY_API_URL>",
"SQUALLY_USE_SYSTEM_CA": "<SQUALLY_USE_SYSTEM_CA>"
}
}
}
}Replace <…> with your own values
Config file
{
"servers": {
"squally": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"squally-mcp"
],
"env": {
"SQUALLY_API_KEY": "<SQUALLY_API_KEY>",
"SQUALLY_API_URL": "<SQUALLY_API_URL>",
"SQUALLY_USE_SYSTEM_CA": "<SQUALLY_USE_SYSTEM_CA>"
}
}
}
}Replace <…> with your own values
Platform CLI
gemini mcp add -e SQUALLY_API_KEY=<SQUALLY_API_KEY> -e SQUALLY_API_URL=<SQUALLY_API_URL> -e SQUALLY_USE_SYSTEM_CA=<SQUALLY_USE_SYSTEM_CA> squally npx -y squally-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 Squally
How do I install Squally in Claude Code?
Run claude mcp add mcp -e SQUALLY_API_KEY=<SQUALLY_API_KEY> -- npx -y squally-mcp. Replace <SQUALLY_API_KEY> with an organization read key.
What does Squally do?
It lets an agent query Squally’s Playwright analytics, including run results, failure details, test stability, and errors grouped across runs. The server is read-only.
What does Squally need to run?
It requires Node 22 or newer, a Squally organization read key, and the Standard plan. A project key for ingesting Playwright results will not work.