Answer prose round 1: 0 🔴, 3 🟡; drop the decision entry and let E and I anchor now

AGENTS.md

@@ -54,7 +54,6 @@ In `docs/decisions.md`:
 - Publish on a version bump
 - Docs describe the release being built
 - Plan items rank by `rank_todo.py`'s goal weights, code health first
-- Until `docs/score-anchors.md` exists, L, E and I are judged by the levels' words
 - Posts announcing a release wait for its publish, and their drafts wait for the finished release
 - No schema validation
 - The gate runs on Deno and Bun
@@ -171,7 +170,8 @@ Reserved for the maintainer whatever any rule here says: changing `version` in `
 bump on `main` publishes, `docs/decisions.md` §Publish on a version bump — every release is the
 maintainer's) and the `NPM_TOKEN` secret.
 
-Before you change a file outside this repo, check whether its directory is a git repo. If it is, change it through a branch and a PR there, as here.
+Before you change a file outside this repo, check whether its directory is a git repo. If it is,
+change it through a branch and a PR there, and merge it only with its owner's approval.
 
 ### Ask, don't guess
 

docs/decisions.md

@@ -601,10 +601,6 @@ weighs 4 times its position weight.
 `todo.md`'s `## Scoring` keeps the formula and the bar, the lowest Score a non-exempt item may have.
 Both are the project's to tune: the maintainer confirmed the bar at 0.1 on 2026-10-21.
 
-## Until `docs/score-anchors.md` exists, L, E and I are judged by the levels' words
-
-2026-10-21, the maintainer, agreeing to track the gap as `todo.md` item 125. An anchor is a shipped item that shows what a level looks like. `score-outcomes` measures only merge commits, and this repo merges by rebase, so it finds no shipped chunk to anchor a level with. L picks a worker's rung, and E + I the reviewers', so scores decide how carefully a change is checked. Serves code health. Valid while `docs/score-anchors.md` does not exist.
-
 ## Posts announcing a release wait for its publish, and their drafts wait for the finished release
 
 2026-10-21, the maintainer, asking for the choice that best reaches the goals. Serves Goal 3.

todo.md

@@ -20,7 +20,7 @@
 | 123 | 0.3.0 |  | defect | **Spell a link title holding a backslash, an entity reference or, in a table cell, a pipe, instead of dropping it.** | 3 | 1 | 2 | 4 | 3 | 7 by 2 | 13.1 |
 | 122 | 0.3.0 |  | defect | **Keep an image's URL and title when a link's text holds the image, instead of dropping them.** | 3 | 1 | 2 | 4 | 5 | 7 by 4 | 10.5 |
 | 95 | 0.3.0 |  | decision | **Find the code-span closers an escape forms, the mark runs a line cannot spell and the text `[` that opens link text by reading the line back, in place of `escapeClosedRuns`, `unspellableRuns` and `lastLinkClose` predicting them.** | 4 | 1 | 3 | 6 | 6 | code health by 7 | 10.2 |
-| 125 | 0.3.0 |  | decision | **Anchor each level of L, E and I in `docs/score-anchors.md` with shipped chunks that `score-outcomes` measures.** | 3 | 2 | 3 | 6 | 4 | code health by 2 | 8.7 |
+| 125 | 0.3.0 |  |  | **Anchor each level of L, E and I in `docs/score-anchors.md` with two or three shipped items.** | 3 | 2 | 3 | 6 | 4 | code health by 2 | 8.7 |
 | 33 | 0.3.0 |  | defect | **Emit a line in time linear in its mark runs, in `adfToLosslessMarkdown` and `adfToPortableMarkdown`.** | 3 | 1 | 2 | 4 | 5 | 9 by 3 | 7.8 |
 | 109 | 0.3.0 |  | defect | **Spell nested lists and quotes in time linear in the markdown.** | 3 | 1 | 2 | 4 | 5 | 9 by 3 | 7.8 |
 | 115 | 0.3.0 |  | defect | **Build text holding no `border` mark in `losslessMarkdownToAdf`, and have `adfToLosslessMarkdown` carry text holding one.** | 2 | 1 | 2 | 3 | 3 | 4 by 3 | 4.6 |
@@ -128,9 +128,13 @@ text. Only the property tests catch drift between these and the reader, and it h
 under the comprehension floor. Reading the line back costs one parse each time a line is tried. The
 run it carries may differ from today's, which changes lossless output (§The formats are API).
 
-### 125. Anchor each level of L, E and I in `docs/score-anchors.md` with shipped chunks that `score-outcomes` measures.
+### 125. Anchor each level of L, E and I in `docs/score-anchors.md` with two or three shipped items.
 
-`docs/score-anchors.md` does not exist, so no level has an anchor. `score-outcomes`, the maintainer's script kept outside this repo, reads merge commits. This repo merges by rebase, which leaves none, so the script finds no finished chunk to anchor a level with. First make the script measure rebase merges. Change it in its own repo, by PR if that directory is a git repo. Then name two or three shipped items for each level, as the scoring rules say. Until this ships, judge each score by its level's words, as `docs/decisions.md` says.
+`docs/score-anchors.md` does not exist, so no level has an anchor. Pick two or three shipped items
+for E and I now, by each level's words and what followed each release, since no tool measures them.
+L anchors need `score-outcomes` first. The maintainer's script, kept outside this repo, reads merge
+commits, and this repo merges by rebase, which leaves none. Change the script in its own repo, as
+AGENTS.md §7 says. Until then, judge each L by its words.
 
 ### 33. Emit a line in time linear in its mark runs, in `adfToLosslessMarkdown` and `adfToPortableMarkdown`.