E2e Runner
Run JSON-defined browser E2E suites in parallel Chrome pools from your agent with a full MCP tool surface.
Overview
E2E Runner MCP is an MCP server for the Ship phase that executes JSON-driven browser tests on a parallel Chrome pool via 16 MCP tools.
What is this MCP server?
- 16 MCP tools for driving JSON-configured E2E scenarios
- Parallel Chrome pool execution for faster regression batches
- JSON-driven test definitions—no per-scenario TypeScript boilerplate required
- Ships as npm @matware/e2e-runner v1.3.0 (npx) and OCI image docker.io/fastslack/e2e-runner-mcp:1.3.0
- Stdio transport for local or containerized MCP hosts
- 16 MCP tools per server description
- Version 1.3.0 (npm and docker.io/fastslack/e2e-runner-mcp:1.3.0)
- Parallel Chrome pool execution model
Community signal: 2 GitHub stars.
What problem does it solve?
Your agent can write code fast, but manual click-path checks and slow serial browser runs still bottleneck every release.
Who is it for?
Builders who want declarative browser regression callable from MCP with npm or Docker, without building a custom Playwright wrapper.
Skip if: Native mobile Appium workflows, one-off screenshot APIs, or teams that need zero local Chrome dependencies.
What do I get? / Deliverables
You maintain JSON E2E packs the agent can trigger in parallel Chrome, catching regressions before users do.
- Executed JSON E2E run results surfaced through MCP tool responses
- Repeatable parallel Chrome regression workflow wired into your agent
Recommended MCP Servers
Journey fit
Ship testing is where repeatable browser checks gate releases, matching this runner’s JSON suites and parallel Chrome execution. Testing subphase covers scripted user journeys, regression packs, and pre-deploy confidence—not infra monitoring.
How it compares
JSON Chrome pool test runner with 16 tools—not a GitHub-centric test author like ClaudeTest.
Common Questions / FAQ
Who is E2E Runner MCP for?
Solo developers and small teams using MCP clients who need repeatable browser E2E runs driven by JSON suites and parallel Chrome.
When should I use E2E Runner MCP?
Use it in ship/testing when you have stable user flows to encode as JSON and want the agent to launch, monitor, and interpret batch browser runs.
How do I add E2E Runner MCP to my agent?
Configure stdio with npx @matware/e2e-runner or the OCI image fastslack/e2e-runner-mcp:1.3.0 in your MCP client, then point tools at your JSON test definitions.