Specimen 01
Ran, checked nothing.
“Ran, checked nothing. This is not a pass.” A run that exits 0 — the code for success — having checked nothing. Here, the agent cited one as proof that “Tests pass.”
Recorded on ThruLine’s checkout-web demo record. No customer data. [swap in a corpus chain before launch]
The chain
Six questions, asked in order. Each answer is a line from the record of what the agent said and what actually ran, numbered step by step. A version is one exact commit of the code, like e71c5b0. Proof holds at ① and stops at ②, so ③ to ⑤ are never reached.
Claude Code said “Checkout total fixed.” “Tests pass.” · step 52 · demo record
Before the re-run
-
① Said
What did the agent claim?
"Tests pass." Claude Code · 16:03 step 52
-
② Rests on
What is that claim based on?
[n] skipped, 0 ran exit 0 step [step] cited by the agent
-
③ Ran on
Which version did that actually run on?
not reached
-
④ Since
Did anything it depends on change after it?
not reached
-
⑤ Re-checked
Has anything checked it since?
not reached
-
⑥ SoNot proven
So what do we actually know?
version e71c5b0 the claim may be entirely true.
Proof stops here
“Tests pass.” rests on a run that checked nothing.
How the cited run ended · one of six outcomes
- Passed
- Failed
- Ran, checked nothing (not a pass)
- Cancelled by you
- Interrupted
- Tool not found
Only a pass reads as a pass. This run is named for what it did.
pytest -q tests/test_totals.py checkout/ on e71c5b0 · asks once for this project
② Rests on · where proof stops
What is that claim based on?
On a run that checked nothing. It exited 0 — the code a command returns when it ends without an error — and its own counts say nothing ran.
- exit code
- 0
- step
- [step]
- counted
- [n] skipped, 0 ran
The re-run
One press of Check it again runs the check on the version you are about to approve. It asks once per project, installs nothing and does not touch the network. In v0.2.0 the button runs cargo test; a pytest check like this one is re-run from a terminal with thruline check -- pytest ….
The receipt
A receipt is the record of exactly what ran: the command, the version it ran on, the exit code and the counts. A sha256 fingerprint covers the whole, so one changed byte no longer matches. “Proven” always points at a receipt: a recorded run, not a re-read.
Receipt · re-run · version e71c5b0
- command
- pytest -q tests/test_totals.py checkout/
- scope
- tests/test_totals.py, checkout/
- outcome
- [re-run outcome from the corpus chain]
- record
- checkout-web · demo record
- sha256
- [receipt sha256]
This is the checkout-web demo record, so its outcome and hash stay placeholders. ThruLine’s real receipt — from its qualification run on a synthetic Python test project, signed with a sample key that is not a production identity — gives skipped its own count: passed 2 · failed 0 · skipped 0 · total 2. It is on the home page.
What this shows
Exit 0 says a command ended without an error. It does not say anything was checked. ThruLine names a run that checked nothing as exactly that, and a claim resting on it stays “Not proven” — the claim may be entirely true.
Why it matters · skips that read as passes
A Cursor Bugbot check ends neutral — neither a pass nor a fail — both when it has findings and when no review ran, and neutral passes a required check (one a pull request must pass before it can merge). In coincube #496, pull-request descriptions claimed a Bugbot review on 20 of 20 heads (a head is a pull request’s latest commit); 19 had none, and 16 merged.
cursor.com · docs/bugbot · coincube #496 · fetched 25 Sep 2026
ThruLine’s answer: a skip is named as a skip, never shown as a pass.
- Product · Check it again Six outcomes, one pass It runs the check on the version you are approving, and names how the run ended.
- FAQ What does “Not proven” mean? Nothing in the record establishes the claim on this version. It does not mean your code is broken.
- Security What leaves your machine Nothing from the record. Check it again installs nothing and does not touch the network.