Blank starter
Begin with your release claims, consequences, evidence references, reproduction steps, and owners.
CSV starter · Markdown starter Direct static files// free launch evidence template
Turn launch claims into evidence you can inspect.
Download a working CSV or Markdown ledger for recording what an AI-built app is supposed to do, what was actually checked, and what still needs attention.
$0 · no account · no email gate · no upload · no saved answer · no tracking beacon
What is included
The complete kit contains a blank ledger in both formats and a worked example for a fictional booking app. Use the example to see how partial or missing evidence changes a claim's status.
Synthetic worked example
Inspect one PASS, REVIEW, CANNOT_ASSESS, and FAIL without mistaking the example for customer evidence.
Example CSV · Example Markdown Fictional app and evidenceFour states, no guessed certainty
The supplied evidence directly supports the claim and the named result was reproduced in the stated environment.
The observed result contradicts the claim or a required control demonstrably failed.
Evidence exists but is incomplete, conflicting, stale, or requires judgment.
Required evidence was not supplied or the relevant environment could not be examined.
What to record for each launch claim
Keep the claim narrow enough to test. Then record the consequence if it is wrong, the exact environment, the evidence reference, what you did, and what you observed.
launch_claimThe behavior or control you expect.consequence_if_wrongWhy the claim matters to this launch.environmentWhere the check was run.evidence_referenceThe file, output, receipt, or record to inspect.reproduction_stepsThe shortest repeatable check.observed_resultWhat happened, without smoothing over gaps.statusPASS, FAIL, REVIEW, or CANNOT_ASSESS.owner and next_actionWho closes the gap and what happens next.Keep sensitive material out of the ledger. Never paste passwords, API keys, production credentials, session cookies, payment information, or private customer data into these files.
Choose the next useful check
The ledger helps you name the gap. These existing paths help you examine it without creating another layer of paperwork.
Not sure what is missing?
Answer six questions without uploading or saving anything.
Run the free launch check →Need to see a finished report?
Inspect a checksum-pinned synthetic example with claims, blockers, and exclusions.
View the sample report →Worried a change will break something?
Map dependencies, inspect test quality, and verify the affected web flow before release.
Change Impact Mapper · Test Quality Auditor · Webapp TesterNeed an independent ledger?
Request a bounded human review for one sanitized app snapshot.
See the $149 review →Launch evidence template questions
What is a launch evidence ledger?
It connects each important release claim to the evidence supplied, the steps used to check it, the observed result, and the next action.
Is this a security audit?
No. The starter is a documentation tool. It does not inspect an app, approve a launch, or certify security, privacy, accessibility, legal, or compliance work.
What should count as launch evidence?
Use inspectable records such as test output, screenshots, logs, checksums, receipts, or a repeatable reproduction. An assistant's statement alone is not evidence.
What is the difference between REVIEW and CANNOT_ASSESS?
Use REVIEW when evidence exists but is incomplete, conflicting, stale, or requires judgment. Use CANNOT_ASSESS when required evidence was not supplied or the relevant environment could not be examined.
Does the starter upload or save project data?
No. These are direct static file downloads. There is no account, email gate, upload, saved answer, or tracking beacon on this page.
Can I use it for a no-code or AI-built app?
Yes. The ledger is based on claims and observable results, so it can be used for coded, no-code, and AI-built applications.
When should I get a human launch evidence review?
A human review is useful when consequential claims remain unresolved, evidence conflicts, or you need an independent claim-to-evidence ledger for one sanitized app snapshot.