Product checks for pump.fun coins
See if a coin's app actually works before you buy.
A robot phone browser uses the project's site like you would (search, open a coin) and screenshots every step as proof.
We write every test by hand. Projects don't write them and can't change the results.
Every $REALRUN buy starts a real test, live. Your buy shows on the run.
Projects tested
Latest run
Loading the latest run
Buys start runs
Products
Each test runs on the real site. Dots show the last 12 runs, newest on the left. Pass counts only count runs of the test as it is written today.
Loading tests
Run log
How to read a result
- Passed
- Every check in that test held on that run. It says nothing about the token, the team, or any part of the site the test didn't touch.
- Failed
- At least one check did not hold. That can be a product bug or a test that went stale after a site change. Open the step and look at the screenshot before you call it either.
- Inconclusive
- The runner couldn't see the result: a blocked page, a timeout, or a bug on our side. It counts as neither a pass nor a fail.
- Fixture
- A run against a local test page we built to check the runner itself. Fixtures never count toward any product.
- The runner never connects a wallet, signs in, sends a transaction or posts anything. It only makes GET and HEAD requests, and only to the sites a test lists.
- Every test is written and reviewed by hand. Nobody can point it at an arbitrary URL.
- We write the tests, not the projects, and nobody can change a result after a run. Each run record is sealed with a hash that shows if it was edited. The hash doesn't prove the record is true; the screenshots and the rerun button are how you check that.