Skip to content

feat(template): a P0 whose scope moves is not a P0 shipped red — 0.10.5 - #37

Merged
indykish merged 2 commits into
mainfrom
feat/template-scope-transfer
Sep 8, 2026
Merged

feat(template): a P0 whose scope moves is not a P0 shipped red — 0.10.5#37
indykish merged 2 commits into
mainfrom
feat/template-scope-transfer

Conversation

@indykish

@indykish indykish commented Sep 8, 2026

Copy link
Copy Markdown
Contributor

Indy: "Spec contradicts a rule → amend spec ... and must be added in the template.md"

The clause was written into one milestone yesterday. This puts it where every milestone reads it.

The gap it closes

The ship gate knew two states for a criterion — met, unmet — plus one escape, a P1 deferral quote. It had no state for scope leaving the milestone. A gate that pretends met and unmet are the only options forces an agent to invent a third, and one did: SCHEDULED, then HANDED OFF — both agent-authored grades, both caught by review, on the same row in the same day.

Deferral ≠ transfer

  • A deferral leaves work unowned inside a closed spec. That is exactly what the P0 gate exists to prevent, and the P1 quote is as far as it goes.
  • A transfer moves the criterion whole: its Dimensions, its verification and its rubric row land in a named successor that carries them as its own P0. Nothing is less owned afterwards — it is owned somewhere else.

MOVED to M{N}_{NNN} R{n} is not ❌, on three conditions that must ALL hold

Each closes a way the word could be abused:

  1. The successor exists and carries the criterion as its own rubric row. A successor that does not carry it is a deferral wearing a new word, and fails as before.
  2. Both specs record the mapping — the closing spec names where each Dimension went, the successor names what it inherited. One-sided assertion is not a transfer.
  3. Discovery carries the owner's verbatim quote. An agent-authored transfer is agent-authored scope reduction.

And a MOVED row is never rendered ✅. The criterion has not been met; it has changed owner, and the rubric says which.

Make

bin/orly update --no-hooks   0 written, 3 already current (19 packs)
make conform                 ALL CHECKS PASSED

Released as 0.10.5 — the clause reaches every consuming repository's docs/TEMPLATE.md on its next orly update.

🤖 Generated with Claude Code

https://claude.ai/code/session_01JLsTMkAD3o53GLoB7rRbyG

indykish and others added 2 commits September 8, 2026 14:03
Indy: "Spec contradicts a rule → amend spec ... and must be added in the
template.md". The clause was written into one milestone yesterday; this puts it
where every milestone reads it.

The gate knew two states for a criterion — met, unmet — plus one escape, a P1
deferral quote. It had no state for scope LEAVING the milestone. A gate that
pretends met and unmet are the only options forces an agent to invent a third,
and one did: `SCHEDULED`, then `HANDED OFF`, both agent-authored grades, both
caught by review, on the same row in the same day.

A deferral and a transfer are different claims. A deferral leaves work unowned
inside a closed spec, which is precisely what the P0 gate exists to prevent —
the P1 quote is as far as that goes. A transfer moves the criterion whole: its
Dimensions, its verification and its rubric row land in a named successor that
carries them as its own P0. Nothing is less owned afterwards.

`MOVED to M{N}_{NNN} R{n}` is therefore not ❌, on three conditions that must
ALL hold, each closing a way the word could be abused:

1. the successor exists and carries the criterion as its own rubric row — a
   successor that does not carry it is a deferral wearing a new word;
2. both specs record the mapping — one-sided assertion is not a transfer;
3. Discovery carries the owner's verbatim quote — an agent-authored transfer is
   agent-authored scope reduction.

And a MOVED row is never rendered ✅. The criterion has not been met; it has
changed owner, and the rubric says which.

Rides the 0.10.4 release rather than taking a version of its own.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01JLsTMkAD3o53GLoB7rRbyG
The scope-transfer clause reaches every consuming repository's docs/TEMPLATE.md
on its next `orly update`.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01JLsTMkAD3o53GLoB7rRbyG
@codecov-commenter

Copy link
Copy Markdown

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

@indykish
indykish merged commit e35dd7e into main Sep 8, 2026
4 checks passed
@indykish
indykish deleted the feat/template-scope-transfer branch September 8, 2026 08:43
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants