box/test
dan-claude-bot c3823e9a3f test: the mint drive mints 'claude-box', the name the template has now
The rebase onto main was textually clean and semantically wrong. #123
renamed the tenant templates (claude -> claude-box, following rig#76) while
this branch was open; the mint-stamp drive was written against the old name
and git had no reason to say so — the two changes never touch the same line.

`box new --template claude` now dies at the template lookup, so all 18 mint
assertions failed on a shimmed mint that never reached a launch line. Renamed
at the two mint invocations and the two stamp assertions that read the new
name back (user.box.template, user.box.role).

Deliberately NOT renamed:

- `user.box.user=claude` — the seed user did not move. #123 split the two
  namespaces on purpose: the template and the rig role are `claude-box`,
  the user the role converges stays `claude`. The assertion pins that split.
- the canned `$STAMPED`/`$PRERENAME` info fixtures — they stand for boxes
  minted BEFORE the rename, which is exactly the case box promises to keep
  reading (cmd_new matches both spellings of user.box.template for the login
  hint). Renaming them would delete the coverage.
2026-07-21 10:51:08 +00:00
..
cli.sh test: the mint drive mints 'claude-box', the name the template has now 2026-07-21 10:51:08 +00:00
labels-reconcile.sh fix(labels): a missing state label must skip the EDIT, not the whole PR 2026-07-20 18:09:07 +00:00
release.sh test: terminate the ci.yml step block at the job boundary too 2026-07-20 20:54:31 +00:00