dual-ai-paper-coach: Command for Claude Code

.claude/commands/iterate.md

iterate is a command for Claude Code from dengxu11111/dual-ai-paper-coach. It costs 18 tokens per session (1,706 once invoked), scanned A, original, MIT.

A command that reads the latest review of a paper and improves the draft in several focused rounds. TDD is not involved here; the review standard mentioned is based on scientific journals such as Science and Nature.

In plain words
What is it for?
Use it to revise a scientific manuscript after an AI review, choosing the number of rounds or the kind of issue to address. It can also trigger another Codex review after the revisions.
Why use it?
It breaks a long revision process into smaller types of changes, such as wording, citations, statistics, limitations, and serious reviewer concerns. This makes it easier to apply review feedback one category at a time.

Command for Claude Code

Written for Claude Code: installed under .claude/. Also seen: mentions Codex.

This is dengxu11111/dual-ai-paper-coach's own configuration. It tells Claude Code how to work on dual-ai-paper-coach itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything dual-ai-paper-coach configures →

Reuse

Borrowing it

Nothing to install: this file belongs to dengxu11111/dual-ai-paper-coach. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/dengxu11111/dual-ai-paper-coach/main/.claude/commands/iterate.md
Clone the repo
git clone --depth 1 https://github.com/dengxu11111/dual-ai-paper-coach

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for iterate

README.md
[![agentmods](https://agentmods.dev/badge/commands/dengxu11111/dual-ai-paper-coach/iterate/github.svg)](https://agentmods.dev/commands/dengxu11111/dual-ai-paper-coach/iterate)
Your own site
<a href="https://agentmods.dev/commands/dengxu11111/dual-ai-paper-coach/iterate"><img src="https://agentmods.dev/badge/commands/dengxu11111/dual-ai-paper-coach/iterate/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for iterate

Your own site · 80×15
<a href="https://agentmods.dev/commands/dengxu11111/dual-ai-paper-coach/iterate"><img src="https://agentmods.dev/badge/commands/dengxu11111/dual-ai-paper-coach/iterate.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 18 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,706 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00018 $0.01706
Opus 5 $0.00009 $0.00853
Sonnet 5 $0.00004 $0.00341
Haiku 4.5 $0.00002 $0.00171

Measured 12d ago against content hash fdeda7ad4956, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

iterate scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/commands/iterate.md · 135 lines

How it starts

The opening of the file, as written. The whole thing — 135 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/iterate — 双 AI 迭代改进

每轮只解决一类问题,跑 N 轮(N 默认 1)。

⚠️ 质量标杆:Science / Nature——Codex 按 S/N 审稿尺度评分。第一轮典型 4-5/10(这是设计上的低分,目的是让你"看见涨分"的真实轨迹)。典型轨迹:4-5 → 6 → 7 → 8(约 4 轮)。

用法

# 推荐起手 ——
/iterate                          # 默认 1 轮(= --rounds 1),按 review 轮次自动选 focus

# 完整迭代 ——
/iterate --rounds 3               # 跑 3 轮
/iterate --rounds 5               # 跑 5 轮(迭代到 8/10)

# 显式控制 ——
/iterate --focus critical         # 只改 review 里 Critical 等级问题(见下方失败排查)
/iterate --focus wording          # 只做措辞修复(= Round 1 默认行为)
/iterate --rounds 1 --focus citations+stats+limits   # 显式指定单轮 focus

--focus 可选值(与"单轮策略"表第二列对齐):wording / citations+stats+limits / critical / mixed。 不传 --focus 时按"轮次 → focus"表自动取(Round 1=wording, Round 2=citations+stats+limits, ...)。

⚠️ 多人协作时同时跑 --rounds 3 会触发 Codex 限流。多人共享账号时建议错峰,或先 --rounds 1 看效果再继续。

单轮策略(关键

每轮只聚焦一类修改,分数才会稳步上升:

轮次 focus 改什么
1 wording overclaim → 弱化措辞(S/N 审稿最严的一项)
2 citations + stats + limits 主源引用(S/N 偏好 primary 不爱 review)、补 Nature Reporting Summary 5 元组(effect size + 95% CI + n + 检验名 + 精确 p + 稳健性)、加 limitations 独立段
3 critical 修 review 里所有 Critical 和 Important,Minor 忽略;补 broad significance framing(Abstract sent. 1 + Intro para 1)
4+ mixed 看上一轮 review 自适应;display items 是否 ≤ 5、refs 是否 ≤ 50

步骤(每轮)

步骤 1:找最新 review

my-paper/review_round_*.md 中最大 N 的那一份。

步骤 2:读 prompts/writer_system.md(守住身份)

防止你在改稿时漂移到"赞美自己"模式。

步骤 3:根据 round number 选 focus

  • Round 1 → focus = wording
  • Round 2 → focus = citations,stats,limits
  • Round 3 → focus = critical
  • Round 4+ → focus = critical(始终聚焦最严重的)

步骤 4:构造改稿 prompt

例如 Round 1(wording):

你正在迭代论文 my-paper/draft.md。

这是 Codex 给你的最新审稿报告:
<贴 my-paper/review_round_<N>.md 的内容>

本轮**只改一类问题**:Wording / Overclaim。
- 把 "proves" → "is consistent with"
- 把 "demonstrates" → "suggests"
- 把 "the first to" → "to our knowledge, among the first to"
- 不要动其他段落

输出:
1. 完整修订版论文 → 写到 my-paper/revision_round_<N>.md
2. 修改对照表(before / after)→ 写到 my-paper/revisions_log_round_<N>.md
3. 把 revision_round_<N>.md 复制到 draft.md(成为新的"最新稿")

Read the full file on GitHub · 135 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 12d ago First seen · 135 lines · 18 tokens per session scan A fdeda7ad4956

Subscribe to this mod's changes

iterate is a command published in the GitHub repository dengxu11111/dual-ai-paper-coach (12 stars, last pushed 3mo ago), licensed MIT. It adds 18 tokens to every session and 1,706 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.