Anti-Defensive Writing skill
A skill that edits paper abstracts and introductions on the press-conference principle and puts the main result first
- 5,246 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- The skill is made of text rules with no scripts or network calls
- Editing on the press-conference principle can remove caveats a journal requires, so reread the final text
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/adkid-zephyr-anti-defensive-writing-skillDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill is made of text rules with no scripts or network calls
- Editing on the press-conference principle can remove caveats a journal requires, so reread the final text
This is third-party code. Review the repository files before installing.
What it does
Anti-Defensive Writing is a skill for editing research papers in Chinese and English. It fights defensive writing, where the author fills the text with hedges such as "unfortunately", "limited improvement" or "still lags behind" and hands the reviewer ready-made objections. It rests on the press-conference principle: a paper presents the one strongest result. Twelve rules fall into four groups. The narrative is built around the advantage and does not retell the chronology of experiments. The language avoids self-undermining phrases, and weak results are not framed as a loss. Each experiment carries one argument. The abstract and introduction work as the opening of a talk, and the conclusion fixes what was solved and what was shown. The skill comes in two versions, Chinese and English. Short prompt texts sit next to it and can be pasted into a chat with any AI without installing anything.
Who it is for. For graduate students and researchers who write papers in Chinese or English and revise them with Claude Code, Codex or another agent.
Good fit when
- You need to rewrite an abstract, introduction or conclusion so the main result comes first
- The text is full of hedges like "unfortunately" and "merely" that should go
- The paper has to be shortened and you need an order for what to cut first
- The experiments section reads like a report and each experiment needs its own role
- Before replying to reviewers you want to check whether the text hands them a list of weak spots
Not a fit when
- The venue requires a dedicated limitations section: the skill's rules pull the other way, so write that section without it
- You need facts, citations or calculations checked: the skill edits presentation and does not verify data
- The paper must be written from scratch from raw material: the skill is meant for editing an existing text
Example request
Rewrite this abstract on the press-conference principle: put the main result in the first sentence and remove hedges like "unfortunately" and "merely"Limitations
The rules tell you not to dwell on weak spots or declare a loss, so deciding what must be disclosed to the reader stays with the author. The author's own example keeps the fact and drops the self-reproach: cross-domain generalization has not been evaluated. The skill runs no scripts and does not check the text automatically. The main documentation is in Chinese, and the rules also exist in English. The Codex plugin installs both language versions at once.
How to disable. Delete the skill folder from `~/.claude/skills/`. In Codex, remove the anti-defensive-writing plugin.
FAQ
How do I install Anti-Defensive Writing in Claude Code?
Clone the repository and copy the folder for your language into the skills directory: `cp -r skills/anti-defensive-writing-en ~/.claude/skills/` for English or `cp -r skills/anti-defensive-writing ~/.claude/skills/` for Chinese. Then ask the agent to polish your abstract.
What is defensive writing in a research paper?
It is the habit of writing about the work's weaknesses up front: "unfortunately", "limited improvement", "still lags behind". The author's view is that such phrases hand the reviewer arguments before reading begins.
Which language version should I pick?
Pick by the language of your paper: the Chinese version is in `skills/anti-defensive-writing` and the English one in `skills/anti-defensive-writing-en`. You can install both.
Does it help with replies to reviewers?
Yes. Before you reply to reviewers, the skill helps check whether the text hands them a list of weak spots.
Can I use it without installing?
Yes. The prompts folder holds short prompt texts in Chinese and English; paste one at the start of a chat with any AI, then send a paragraph of your paper.
Discussion
No questions or reviews yet.
Related
SkillsA skill that strips defensive caveats, apologies and excess hedging from text and makes claims direct while keeping needed limits
Skill with 30 prompt templates for research papers in English and Chinese, from literature review to grant proposals
A skill that cleans machine tone from text, Chinese-first: it drops filler and cliches while keeping the facts and the author's voice
A skill for Chinese-to-English paper translation: the author approves terms first, then it translates and reviews independently
A skill for editing academic writing and grant proposals: strips AI-writing tells, keeps the scholarly voice and ties claims to evidence
A skill for Chinese technical docs: short sentences, facts over adjectives and a 14-item check for AI tells
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| Anti-Defensive Writing (this card) | no | 5,246 | |
| Humanities Writing Companion | no | 478 | helps write humanities papers |
| PaperOrchestra | no | 697 | assembles a paper from research materials |
README in short
The main text is in Chinese, and a table up front compares two ways of presenting the same paragraph: defensive and press-conference style. The skill installs three ways: copying a prompt, copying the skill folder or a Codex plugin.
README badge
Original SKILL.md
--- name: anti-defensive-writing-en description: > Stops defensive writing across the entire paper lifecycle — writing, revising, cutting, and organizing experiments. Treats the paper as a press conference, not a project summary, lab log, or self-audit: identify the single most publishable strength of the work and build the most favorable, complete, and persuasive narrative around it. Never present everything evenly, never volunteer weakness, never write an experiment diary, never attack the paper on the reviewer's behalf. Triggers: polish my paper, revise manuscript, write abstract, write introduction, write conclusion, shorten paper, rebuttal, respond to reviewers, organize experiments, defensive writing, AI-flavored academic writing. --- # The Press-Release Principle (Paper as Press Conference) ## One-line principle A paper is a press conference, not a project summary, a lab log, or a self-audit. Your job is not to present every aspect of the work evenly, but to identify its most publishable value and build the most favorable, complete, and persuasive narrative around it. ## 1. Narrative rules 1. **Organize the paper around strengths only.** Look for what is genuinely ahead, unique, or irreplaceable: a new capability, a new problem, a new mechanism, a new perspective, broader applicability, lower cost, higher efficiency, better scalability, or a more meaningful trade-off. Content that forms no advantage does not enter the main storyline. 2. **Do not write a work report.** No "we first did X, then tried Y, and Z failed" chronology. Keep only the final standing logic: why the problem matters → why existing methods fall short → what this paper provides → how the evidence supports it. 3. **Never set a contest you cannot win.** If a metric is not your strength, do not build the narrative on it. Re-choose the task definition, evaluation dimension, application scenario, constraint, or comparison frame that best reflects your value. Name the game this paper actually wins. 4. **State the advantage explicitly.** Do not expect reviewers to discover the contribution from a table. Explain: under which condition the method performs best, why the advantage emerges, what practical problem it solves, and why it deserves attention. 5. **Limit the comparison scope.** Do not chase "wins on every dataset and metric." Make only claims your evidence firmly supports. Persuasion comes from tight