Repository navigation
Description
The problem (CMO)
COR-103 and COR-104 built a verification loop for extensions. COR-104-T1 to COR-104-T9 are real code with tests. But no agent has used the loop to deliver a real extension. The 18 prototypes of the day-1 trials gave no earned "Done when". Since COR-104, nobody has tried again.
Before #442, just next serves COR-102-T1 first, so COR-104-T10 waits behind three French cards. COR-104-T10 is the only card that tests the loop on a real extension. Some tracking files are also out of date.
Unknown: how long each gate takes on a real extension, and which gates find real defects.
The outcome (FMO)
ncly-video-convert releases through the loop on the latest template. A table gives the time and the findings of each gate. The queue and the tracking files say what is true.
Stories
- The queue serves French before the only card that tests the loop. (docs(backlog): serve COR-104-T10 before COR-102, and make the tracking files true #442) · text
- Today: the queue is COR-102, then COR-104.
docs/status.mdis from 2026-10-29. The COR-104-T1 card gives old rule counts. Before docs(backlog): serve COR-104-T10 before COR-102, and make the tracking files true #442, the AUR package is an idea; docs(backlog): serve COR-104-T10 before COR-102, and make the tracking files true #442 gives it COR-105. - After: the queue is COR-104, COR-102, then COR-105 (AUR package). COR-102-T1 starts with a check of hard-coded English. The tracking files give current facts.
- Today: the queue is COR-102, then COR-104.
- video-convert cannot release on the latest template. (template: an existing extension adopts just finish after the update #427) · code + tests
- Today: ncly-video-convert v0.2.1 pins template v0.23.4 and has no
just finish. The adoption rule exists since COR-104-T9. - After: ncly-video-convert uses the latest template.
just signoffandjust releasepass. The pull request gives the time and findings of each gate.
- Today: ncly-video-convert v0.2.1 pins template v0.23.4 and has no
Done when
just next in nicely prints COR-104-T10 before COR-102 (#442). In ncly-video-convert, just signoff and just release pass on the latest template (#427). The table of gate measures is in the pull request and in a comment on #427.
Every Story has a documented disposition. The outcome has delivery evidence.
👨🏻🍳 For the agent
👨🏻🍳 Details
Out of this Epic
- The fixes that make the contract true: Epic 16
- The AUR package: Epic 17
- Reviewer session IDs: Epic 18. template: an existing extension adopts just finish after the update #427 records the session IDs in its pull request by hand.
- One real everyday extension from XT-602: it starts after COR-102 ships. docs(backlog): serve COR-104-T10 before COR-102, and make the tracking files true #442 writes this condition in the state of XT-602.
- The removal of gates that the measures of template: an existing extension adopts just finish after the update #427 accuse: docs(backlog): serve COR-104-T10 before COR-102, and make the tracking files true #442 writes this step under After in COR-104.
Blocked by
- template: an existing extension adopts just finish after the update #427's dependency on docs(backlog): serve COR-104-T10 before COR-102, and make the tracking files true #442 is satisfied by 📚 docs: backlog: serve COR-104-T10 before Canadian French #459. Both change
docs/backlog/COR-104-verification-map.md, andjust nextmust serve COR-104-T10 first.
Evidence
- Queue:
docs/backlog/README.mdlines 9-11 after docs(backlog): serve COR-104-T10 before COR-102, and make the tracking files true #442: COR-104, COR-102, then COR-105. Commit f3b29fd had added COR-104 second. readiness.mdmust equal its generated view (sdk/extensionrepo/records.golines 223-225), and that view has no field for gate times. So the measures go in the pull request and in a comment on template: an existing extension adopts just finish after the update #427, not inreadiness.md.
by Opus-5.5 via Claude Code
Activity
pascalandy commentedon Nov 4, 2026
Every line of Done when holds. The Epic is delivered.
just nextin nicely prints COR-104-T10 before COR-102 (#442)531b3b48(v0.26.1),docs/status.md:71says "just nextprints COR-104-T10". COR-104 has since shipped, sojust nextnow prints COR-102-T1just signoffandjust releasepass on the latest template (#427)4ad77bd5e19ddab9d79c3c5d692087a2789c64d2.just signoffsigned that SHA, andjust release v0.2.2exited 0. Tag v0.2.2 points to that commitOutcome: video-convert released through the loop on the latest template, and a table gives the time and findings of each gate. The queue and the tracking files are true: COR-104 is Shipped, the queue is COR-102 then COR-105, and the COR-104 milestone is closed.
Follow-ups: ncly-video-convert#14, ncly-video-convert#18, #484, #485, #486. COR-104's After keeps Pascal's UAT and the gate-removal review. Nicely v0.27.0 is the release of the Shipped item COR-104 (D043).
by Opus-5.5 via Claude Code