Whatever you're stuck on, it's in here
Watch an actor work through a real site before you sign up. Follow a tutorial to your first checklist. Look up any part in the docs when you need it.
Start here
Three ways in, depending on how you like to learn.
Tutorials
Step by step, from your first actor to a checklist that runs on every deploy.
- Build an actor that behaves like your audienceInstructions and knowledge that make an actor feel real.
- Write checklists that catch real regressionsStructure checks around outcomes and read the evidence.
- Run a user interviewSections, items, recruitment, and reading the report.
- Automate multi-step workflowsChain checklists, conditions, and PR reviews together.
- Turn runs into insightsRead results, screenshots, recordings, and insights.
- Collect stakeholder feedback on a prototypeInstall the feedback widget and triage what reviewers pin.
- Review designs from SlackShare a link in Slack and get a reviewed screen back in the thread.
- Test a product behind your firewallRun actors against an internal product with a worker.
- Automate Stunt Double from ClaudeConnect the MCP server and drive your workspace from Claude.
Guides
How Stunt Double fits into Figma, Claude, your codebase and your review process.
- Integration · 4 minDesign in Claude, validate with personasDesign an artifact in Claude, publish it, and test it with AI personas without leaving the conversation.
- Integration · 5 minTest Figma prototypes with AI personasWalk prototypes with personas and re-check the built flows automatically when a file changes.
- Integration · 4 minThe fastest idea-to-evidence loop with Figma MakePublish a Make site and have AI personas try to use it before an engineer touches it.
- Integration · 5 minVerify every change with Claude Code, Cursor, and GitHubWire Stunt Double into your coding agent so every preview deployment is walked by a persona before merge.
- Playbook · 6 minCollect comments straight from your siteInstall the feedback widget so visitors pin comments in context, then triage them alongside what your actors find.
- Playbook · 7 minContinuous validation: brand, compliance, continuity, design systemTurn brand, tone, legal, continuity, and design-system standards into checks that run on every deploy.
For developers and agents
Drive Stunt Double from code, from an MCP client, or from your own machine.
- API overviewHow to drive Stunt Double from code, including auth and rate limits.
- MCP serverRun checklists and interviews from Claude, Cursor or any MCP client.
- MCP on GitHub (opens in new tab)The public repository for the Stunt Double MCP server.
- Mac app (download for macOS)Run actors in your own browser, on your machine and your network.
- llms.txtThe site in one plain-text file, for coding agents and assistants.
Keep up
What we have shipped, what we are learning, and where Stunt Double fits.
- ChangelogEverything we have shipped, newest first.
- Backstage blogNotes from building and using Stunt Double.
- Use casesOnboarding, checkout, AI-built apps, pull requests, agent readiness and more.
- GlossarySynthetic users, agent experience, UX regression testing and more, defined.
- AlternativesWhere Stunt Double fits beside panels, scripts, QA and analytics.
- SecurityHow we handle your credentials and data.
Get started
Your cast is ready to start
Create your first actor in minutes.