A green test is not evidence
Tested, packaged and observed are three different claims. Most arguments about what a system can do come from treating them as one.
| Claim | What it establishes | What it does not |
|---|---|---|
| Tested | Code exists and a test exercises it, including its failure path. | That anyone has seen it work in the product. |
| Packaged | It is inside a build that was produced and signed. | That it was ever opened or observed. |
| Observed | Someone ran the application and watched it happen. | That it is covered by a test, or that it will happen again. |
Applied to the current version, the three words split the same release cleanly: the workbook engine is tested; three of four interface fixes are packaged and unobserved; the preflight refusal was observed in a live run.
#Why the distinction is load-bearing
A spreadsheet was found in a data room and looked like proof that the engine worked. It had been written by a Python agent in an earlier version, before the engine existed. A file on disk is not evidence until you know what wrote it.
The same discipline runs through this site. Where a page describes a capability that is written and tested but never watched, it says so in those words, and it says what would close the gap.
#Sources
- Deskworth OS — handover report 0.0.22DeskworthInternal document, not published
Read next
- MethodologyHow a Deskworth run is set up so that its result can be argued with: scope, as-of date, admissible sources, missing data, provenance, checks, and what counts as evidence.
- Field notesShort notes from building the thing: one observation, its cause, the fix, the test that holds it, and what is still open.