tighten-paperlisted
Install: claude install-skill zernie/paperlint
<!-- vigiles:sha256:79a018d2fd2bf3e7 compiled from skills/tighten-paper/SKILL.md.spec.ts -->
# tighten-paper — the developmental edit
## Run me
🔴 FIRST, before any other step:
```
node .claude/skills/paper-pipeline/scripts/announce.mjs tighten-paper <paper-dir>
```
An advisory pass cannot be observed failing — silence is both its error state and its normal
state — so starting is an event, and events get written down.
Sentence-level passes make a wall of text into readable text. They do **not** tell you the wall shouldn't be there
at all. This pass is the editor who reads the whole paper once and asks: *is this the right paper, at the right
length, delivering the right point?* Its deliverable is a **cut/fold/merge/reorder plan and a length verdict**, not
reworded sentences.
**Run it as a persona subagent too** (see `paper-pipeline/references/writing-craft.md` — the persona mechanic): an
editor persona who reads the whole thing once, fast, the way a busy reviewer skims before deciding. A frontier model
grading section-by-section never feels the *cumulative* drag; a persona reading cover-to-cover does.
## 🔴 Run this ONCE, at the end — never after each addition
The failure this rule exists for is not hypothetical; it is what produced the worst score
`the reference paper` ever received. The body sat exactly at its 8-page limit, so every content
addition pushed it to 9, and each time prose came out to get back — around fifty times, sixty
characters at a go, always inside