We provide proactive agents that work on your projects in the background. Learns your personality. Remembers and keep track. Always with your phone or Slack.
Malmö, Sweden
Joined November 2022
- Tweets217
- Following22
- Followers339
- Likes141
Your agents can now easily create both music and video. Like this awesome epic metal song: squidler.io/blog/agents-make…
We now support hosting your apps on your own domain. Create the app in Squidler and host it on your public hostname.
squidler.io/blog/apps-on-you…
We've given our agents voices now. You talk to them, and have them respond. Or just ask them to read something for you. In multiple languages.
squidler.io/blog/agents-read…
We've extended the apps you can create, so now it's easy for you to setup multiple users, and invite anyone to use your apps behind a login window.
squidler.io/blog/apps-with-u…
We just shipped full agent EU residency.
squidler.io/blog/data-reside…
Mattias Ask | Squidler | Your taste is Impeccable, Your AI Is Generic | ... youtu.be/bhAAOa8sVi8?si=d1vJ… via @YouTube
replit agents are great until feature 5 quietly breaks feature 2. squidler's mcp runs your real user flows every iteration so the agent stops playing regression whack-a-mole and actually moves forward
replit.com/partners/squidler
Your e2e tests as plain-english markdown, committed next to the code they exercise. Reviewed in the same PR, versioned with the app, never drifting out of sync.
Same files run locally while you build and in CI on every pull request. One source of truth.
qa.squidler.io/product-updat…
Squidler retweeted
Replit Agent builds your app. Squidler tests it like a real user. Replit Agent fixes what's broken.
That's the full AI QA loop, and it's now live in Replit's MCP library. You describe what your app should do in plain English. Squidler navigates it the way a real person would. Issues flow back automatically and get fixed. No test-writing skills required.
Build with Replit. Test with Squidler. Ship with confidence.
Official today: Squidler is in @Replit MCP library. Build with #Replit. Test with Squidler. Replit Agent builds, Squidler runs real user flows against the live URL, results loop back to the agent. No selectors, no scripts. replit.com/partners/squidler
Official today: Squidler is in @Replit MCP library. Build with #Replit. Test with Squidler. Replit Agent builds, Squidler runs real user flows against the live URL, results loop back to the agent. No selectors, no scripts. replit.com/partners/squidler
Most teams testing AI-generated code start from the code. Reading it, trying to understand what it does, then writing tests for that. Problem is you're testing the implementation, not the intent. Better to start from what the user should experience and work backwards. If the test reads like a user story you probably got it right
The hardest part of testing AI-built apps isnt writing the tests. Its knowing what to test when you didnt write the code yourself. Natural language tests help because they force you to describe what the app should do from the users perspective. Not implementation details you dont understand anyway.
Every AI coding agent hits the same wall. Test fails, agent retries, same failure, loop until timeout. The problem isnt the agent - its the test. An assertion like 'button exists' tells you nothing about what went wrong. Write tests that describe what the user experiences and suddenly your agent can actually fix things.
Hot take: if your AI coding tool cant verify its own output, its just a fast typist. We plugged Squidler into Cursor and Claude Code via MCP. Now the agent writes the feature, then runs behavioral tests against it before you even look at it. The loop closes itself.
Most test suites test implementation. Click this button, check this element, verify this class. Then someone renames a div and half your tests break. We write tests that describe what the user should experience. No selectors, no scripts. The test survives every refactor because it never knew about your code in the first place.
GPT-5.5 just dropped and everyone's testing prompts. Nobody's testing the code it writes. Your shiny new feature works in the demo. Does it work when a user hits it sideways? Plain English test specs, real browser, real feedback loop.
We just shipped localhost testing. Your AI coding assistant (Claude Code, Cursor, Codex) can now run your Squidler test cases against your local dev server - before you commit. One npm package, zero API keys, real test quality. The feedback loop just got a lot tighter."
Link: squidler.io/product-updates/…