for Replit Agent builders

Replit Agent ships your app. Real humans make sure it works.

Agent-built code passes the unit tests but breaks on the login button. TestMyVibes puts a verified human tester on your Replit app after every Publish — full screen recording, structured bug list, suggested fixes the Agent can apply on the next iteration.

Test my Replit app → See how it works

5 free credits · No card required · Same-day reports

Why human QA

Why Replit-built apps need a human QA layer

Replit Agent is brilliant at scaffolding and refactoring. It is bad at noticing that the OAuth callback redirects to a 404, that the mobile checkout button is hidden under the keyboard, or that pressing Back twice logs the user out. That is what humans catch.

Agent doesn't click the buttons

Replit Agent reads code, not screens. It cannot tell that the form submits but the success toast never appears. A human tester clicks every button and reports what actually happens.

Multi-step flows break silently

Sign up → verify email → onboarding → first action. Each step works in isolation. The chain breaks. A human walks the whole flow end-to-end on a real Replit deployment.

Mobile is where Agent code dies

Replit previews are desktop-first. Your users are on phones. Our checkers test on real iOS and Android devices, not browser dev tools.

The Agent can fix what we report

Every bug we file ships with a suggested fix. Hand it back to Replit Agent and it patches the code on the next iteration — that is the auto-test loop.

Real human checkers testing Replit-built apps across desktop and mobile browsers
How to set it up

How to wire TestMyVibes into Replit

Two minutes. Tell Replit Agent to install our skill file, paste your API key, and every Publish triggers a human QA pass.

  1. Get your TestMyVibes API key

    Sign up (5 free credits, no card) and copy your key from Dashboard → API Keys.

  2. Tell Replit Agent to install the skill

    Prompt: "Install the TestMyVibes auto-test loop skill from testmyvibes.com/agent-skills/replit and store my API key as a Replit Secret called TMV_API_KEY."

  3. Run the loop

    Prompt: "Use the testmyvibes-loop skill to test and fix my app." The Agent publishes, calls our API, waits for the human report, applies the suggested fixes, republishes, and repeats — capped by your iteration and credit budget.

  4. Watch the report in your dashboard

    Each test gives you a screen recording, pass/fail checklist, severity-rated bug list, and a health score. Plug the deploy hook into Replit Deployments and it runs on every Publish automatically.

.local/skills/testmyvibes-loop/SKILL.md Download SKILL.md 
# 1. Save the rule file
#    Download from https://testmyvibes.com/agent-skills/replit
#    Save this file as `.

# 2. Set your TestMyVibes API key as a secret named TMV_API_KEY.

# 3. Prompt the agent:
#    "Run the TestMyVibes auto-test loop."

# Under the hood, the agent calls:
curl -X POST https://testmyvibes.com/v1/replit/test \
  -H "Authorization: Bearer $TMV_API_KEY" \
  -H "Content-Type: application/json" \
  -H "X-Agent-Skill: replit" \
  -H "X-Agent-Iteration: 1" \
  -d '{ "url": "https://your-app.example.com",
        "feature": "what you just built" }'

Replit Agent skill — paste once, then run the loop on every deploy.

Looking for the deeper API integration? See the Replit integration page →

Proof

What we catch on Replit Agent apps

Login redirect loops Agent wired the OAuth callback to the wrong route — user bounces between /auth and /.
Stripe checkout 404 Agent referenced an env var that wasn't set in Deployments. Local works, production doesn't.
Mobile keyboard covers Submit Form is fixed-position on mobile. Tester can't tap Submit without dismissing the keyboard.
Empty-state crash Dashboard renders fine with seed data, throws when a brand-new user with zero records lands on it.
  • Verified human checkers — every tester passes a background QA test before joining.
  • Real browsers on real devices (desktop, mobile, tablet) — no headless emulation.
  • Full screen recording of every session — replay exactly what your tester saw.
  • Average response time under 30 minutes during business hours.
Pricing

$1 per credit. 5 free to start.

1 credit = 2 minutes of human testing = $1.00. You only pay for time used. 5 free credits when you sign up — no card required. Compare that to one angry-user support ticket.

See full pricing →
FAQ

Replit + TestMyVibes FAQ

Does this work with Replit Deployments?

Yes. Add a deploy hook URL from your TestMyVibes dashboard to Replit Deployments and a human test fires after every successful publish.

Can the Agent apply the fixes automatically?

Yes — that is the entire point of the auto-test loop skill. The Agent reads our suggested fixes and patches the code, then republishes and re-tests until clean (capped by your budget).

How is this different from running my own tests in Replit?

Unit tests confirm code compiles. Humans confirm the app actually works for real users on real devices. Both belong in your pipeline.

Will my API key be exposed?

No. The skill stores it as a Replit Secret. It is never committed to your Repl.

Stop shipping Agent code blind.

5 free credits. No card. First report typically in under 30 minutes.

Test my Replit app →