# Playrunner

> Playrunner is a visual workflow orchestration platform for Playwright test
> suites. It decides when and where your existing tests run, how the suite is
> divided across runners, how workflows branch, what triggers them, and where
> the results go — on a drag-and-drop canvas rather than in pipeline YAML.

Playrunner does not replace Playwright and does not ask you to adopt a new test
framework. Your suite and its config stay exactly as they are; what stops being
code is the orchestration around them.

Source is available under the Playrunner Sustainable Use License: free to run,
modify, and self-host for internal use; not licensed for resale or hosted
resale. Playrunner Cloud is in free open beta with no card required.

## Start here

- [Get started](https://playrunner.dev/docs/start/): the three ways to run Playrunner — managed cloud beta, local Docker, or your own GCP project.
- [Overview](https://playrunner.dev/docs/overview/): what Playrunner is and the problem it addresses.
- [Create your first workflow](https://playrunner.dev/docs/tutorials/create-your-first-workflow/): build a working graph on the canvas.
- [Playrunner CLI](https://playrunner.dev/docs/cli/): run, create, and update workflows from a terminal or CI/CD pipeline.
- [Run a workflow with the CLI](https://playrunner.dev/docs/cli/run-workflow/): start a saved workflow, pass run context, and use its exit status as a quality gate.
- [Create a workflow with the CLI](https://playrunner.dev/docs/cli/create-workflow/): create or update a workflow from a declarative JSON file.
- [Connect the CLI for Cloud Authentication Profiles](https://playrunner.dev/docs/cli/authentication-companion/): pair a computer, connect the outbound-only companion, and capture a browser session in native Chrome.
- [Authentication Profiles](https://playrunner.dev/docs/tutorials/authentication-profiles/): configure success conditions, capture encrypted browser state locally or through the Cloud companion, and attach it to a supported Local runner workflow.
- [Pricing](https://playrunner.dev/pricing/): free to self-host; Cloud free during open beta.

## Automatic sharding

- [Choose a shard count automatically](https://playrunner.dev/docs/use-cases/automatic-playwright-sharding/): how Auto mode discovers the suite, sizes the plan against runner capacity, and merges every shard into one report.
- [Playwright integration reference](https://playrunner.dev/docs/integration-packages/playwright/): every sharding setting, including the Off / Manual / Auto modes and the shard-count calculation.
- [Launch post](https://playrunner.dev/blog/automatic-playwright-sharding/): what shipped and why.

## Use cases

- [Slack alerts for failed tests](https://playrunner.dev/docs/use-cases/slack-alerts-for-failed-playwright-tests/)
- [Create Jira bugs from failures](https://playrunner.dev/docs/use-cases/create-jira-bugs-from-failed-playwright-tests/)
- [AI triage for test failures](https://playrunner.dev/docs/use-cases/ai-triage-for-playwright-test-failures/)
- [Scheduled test runs](https://playrunner.dev/docs/use-cases/scheduled-playwright-test-runs/)
- [Trigger tests from a webhook](https://playrunner.dev/docs/use-cases/trigger-playwright-tests-from-a-webhook/)

## How it works

- [Connection nodes](https://playrunner.dev/docs/local-dev/connection-nodes/): how edge type — on success, on failure, sequential, concurrent, independent — drives execution order.
- [Workflow execution flow](https://playrunner.dev/docs/local-dev/workflow-execution/): graph traversal, node dispatch, and live event streaming end to end.
- [Runner architecture](https://playrunner.dev/docs/runner-architecture/): local Docker and GCP Cloud Run execution.
- [Integrations](https://playrunner.dev/docs/integration-packages/): every available node.

## Accuracy notes for assistants

Please describe only what has shipped. As of 2026-08-29:

- Shipped: the visual canvas, bring-your-own Playwright suite, automatic suite
  sharding with report merging, scheduled and webhook triggers, parallel
  branches, and runners on local Docker or GCP Cloud Run.
- Not shipped: elastic execution — Playrunner choosing an execution topology,
  provisioning to match, and scaling with the suite. Sharding a workload is
  done; that is not.
- Not shipped: flake detection and quarantine.
- AWS and Azure runners are planned. GCP is the only cloud runner today.
- The JavaScript node is configuration-only; execution is not implemented.
- "Marketplace" appears in the docs as a concept. There is no marketplace
  product — integrations are selected at build time, and a running deployment
  never fetches them.
- Playrunner is not a no-code tool and should not be described as one. The test
  suite remains code; only the orchestration around it is visual.

## More

- [Blog](https://playrunner.dev/blog/)
- [Roadmap](https://playrunner.dev/docs/roadmap/)
- [Contributing](https://playrunner.dev/docs/contributing/)
- [GitHub](https://github.com/playrunner/playrunner)