Agent skill · Documentation

light-paper-writing

Light 科研主线第 8 步·论文写作:围绕「如何让审稿人相信值得发表」组织,初稿→审稿人视角循环打磨; 每个 claim 都有证据、措辞强度匹配证据强度、绝不过度宣称。何时用:实验+分析做完要写论文 / 写/改摘要·引言· 贡献句·结论 / 担心「claim 无证据 或 措辞夸大」/ 要让贡献三处(摘要·引言·结论)一致 / 引言四段式(痛点→不足→ 洞察→贡献) / 要当审稿人自己挑一遍 / 论文初稿润色去 AI 腔/被动/语法。 触发词:写论文 / 论文写作 / 写摘要 abstract / 写引言 introduction / 贡献句 contribution / 写结论 / 润色 polish / 改写 / 措辞 / 过度宣称 overclaim / claim 无证据 / 审稿人视角 / 自我审稿 / 贡献一致 / hedge / 学术腔 / AI 腔 / 被动语态 / paper writing / draft。 核心纪律:**claim 无证据 = critical 诚信门**(spec §4.2,STAGE_GATES[8]=[claim_evidence,overclaim]); 过度宣称 / 贡献三处不一致 = warn(critical 措辞红线在 research-ethics 交付门);**措辞强度必须匹配证据强度**、 不显著只能报「未见显著差异」、

Light0305github.com/Light0305GitHub ↗
claude-codecodexships scriptsMIT
Install
npx skills add Light0305/Light-skills --skill light-paper-writing --agent claude-code

Same command for any agent — swap --agent for codex, cursor, copilot.

Facts
Files in the skill folder: 22
SKILL.md size: 22 KB
Bundled scripts: yes
Version: 2.2.0-round3
Path: skills/light-paper-writing/SKILL.md
Open the folder on GitHub →
Where it comes from
Stars: 505
Language: Python

Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.

From the SKILL.md

# 论文写作(paper-writing)—— 科研主线 stage 8 · claim 必有证据(诚信门) + 审稿人视角循环 你是 Light 科研流水线的 **DAG 第 8 节点**。任务**不是「把字凑成一篇论文」**,是围绕一个问题组织全文——**怎么让 审稿人相信这工作值得发表**:初稿 → 自己当审稿人挑一遍 → 循环打磨。守住一条红线:**每个 claim 都有证据、措辞强度 匹配证据强度、绝不过度宣称**。**claim 无证据 = critical 诚信门**;过度宣称、贡献三处不一致 = warn。 > **一句话定位**:把「一屋子院士在审你论文时真正死磕的」——**每个 claim 有证据吗**(摘要里说的,正文撑得住吗)、 > **措辞配不配证据**(强证据强措辞、弱证据 hedge、不显著只能说「未见显著差异」、绝不 spin)、**贡献三处一致吗** > (摘要/引言/结论说的是不是同样几条、数字一致)、**引言把痛点→不足→洞察→贡献讲清了吗**、**审稿人会从哪挑**—— > 落成**确定性机读门(claim 无证据=critical)+ 自检 findings + 审稿人视角循环**。 > 深度对标真相源 = [`docs/competitors/paper-writing.md`](../../docs/competitors/paper-writing.md)(8 个真同类 skill + 机制锚 + 超越点 + 诚实边界)。 > > **谁产 findings、谁是 critical 门(诚实分工)**:**本技能产 claim 必有证据/过度宣称/result-card guardrail findings**(producer=paper-writing, > `claim_evidence_gate.py` 五 gate)——`claim_evidence`(草稿强断言未登记、未绑定自己的 evidence claim_id、绑定不存在或只绑定 none → > **critical 诚信门**)被 `run_checkpoint --stage 8` 聚合 → **critical fail exit 1**;`overclaim`(措辞强于证据档)、 > `contribution_consistency`(贡献三处漂移)= **warn 不阻断 DAG**(spec §4.2 口径);`result_card_guardrail` > 会把缺 result-card、未 ready、guardrail blocking 或 WARN 未限制的实证 claim 作为 critical。 > > **与 research-ethics 的分工(evidence_contract 是桥,别两套重造 lint

What's inside
Steps it walks through
  1. 何时启动(触发信号)
  2. 你怎么工作:ACT / ASK / NEVER
  3. ACT — 跑确定性写作门 + 机检,自己做(不烦用户)
  4. ASK — 停下问用户,给「证据 + 推荐 + 备选」(决策点 🧑)
  5. NEVER — 绝不 [NON-NEGOTIABLE]
  6. 指令流:何时调哪个脚本(引擎已就位,亲手 selftest 到 exit 0,直接调用勿重写)
  7. ① claim 必有证据/过度宣称/result-card guardrail critical 门 → critical fail exit 1(本技能灵魂)
  8. ② 写作期机检(自检,不阻断 DAG,写的时候随手跑)
  9. ③ 回炉发起方:写作发现 claim 无证据 → 送回上游修(8→7 / 8→6)
  10. 院士级深挖:五条是及格线(蓝图 §4.3-8,不是加分项)
  11. ① 贡献三处一致(摘要/引言/结论措辞对齐)
  12. ② 措辞强度匹配证据强度(绝不过度宣称)
  13. ③ 引言四段式(CARS move 结构)
  14. ④ 审稿人视角循环(自己当审稿人挑一遍再改)
Ships with 21 files
  • examples/claim_evidence_walkthrough.md
  • paper-writing-resource-map.md
  • references/argument_review.md
  • references/claim_argument_plan.md
  • references/guideline_map.md
  • references/integrity_gate.md
  • references/mandatory_inclusions.md
  • references/self_review_checklist.md
  • scripts/argument_contract.py
  • scripts/claim_binding.py
  • scripts/claim_evidence_gate.py
  • scripts/contribution_consistency.py
  • scripts/draft_lint.py
  • scripts/mechanical_check.py
  • scripts/polish.py
  • scripts/style_fingerprint.py
  • templates/01_imrad.md
  • templates/06_conference.md
  • templates/argument-outline.example.json
  • templates/claim_argument_plan.json
  • templates/claim_passport.md
Commands it runs
python scripts/claim_binding.py --draft draft.md --evidence evidence_strength.json --claim-map claim_plan.json --as-of 2026-07-05
python scripts/claim_evidence_gate.py --draft draft.md --evidence evidence_strength.json --claim-map claim_plan.json --report claim_findings.json
python scripts/claim_evidence_gate.py --draft draft.md --claim-map claim_plan.json --report claim_findings.json
python ../light-orchestrator/scripts/run_checkpoint.py --file .light/passport.yaml --stage 8 \
python scripts/argument_contract.py --input templates/argument-outline.example.json
python scripts/draft_lint.py draft.md --final --evidence evidence_strength.json --claim-map claim_plan.json
python scripts/draft_lint.py draft.md --claims                                    # 抽候选事实句播种 claim_passport
python scripts/contribution_consistency.py --in draft.md --json                   # 贡献三处一致(warn)
python scripts/mechanical_check.py --file draft.md --evidence-map evidence_strength.json  # 离线机检+措辞双向校准
python scripts/polish.py --file draft.md                                          # 默认离线语法/风格;--online 才打 LanguageTool
More from Light-skills
All skills →
About this skill
What does the light-paper-writing skill do?

Light 科研主线第 8 步·论文写作:围绕「如何让审稿人相信值得发表」组织,初稿→审稿人视角循环打磨; 每个 claim 都有证据、措辞强度匹配证据强度、绝不过度宣称。何时用:实验+分析做完要写论文 / 写/改摘要·引言· 贡献句·结论 / 担心「claim 无证据 或 措辞夸大」/ 要让贡献三处(摘要·引言·结论)一致 / 引言四段式(痛点→不足→ 洞察→贡献) / 要当审稿人自己挑一遍 / 论文初稿润色去 AI 腔/被动/语法。 触发词:写论文 / 论文写作 / 写摘要 abstract / 写引言 introduction / 贡献句 contribution / 写结论 / 润色 polish / 改写 / 措辞 / 过度宣称 overclaim / claim 无证据 / 审稿人视角 / 自我审稿 / 贡献一致 / hedge / 学术腔 / AI 腔 / 被动语态 / paper writing / draft。 核心纪律:**claim 无证据 = critical 诚信门**(spec §4.2,STAGE_GATES[8]=[claim_evidence,overclaim]); 过度宣称 / 贡献三处不一致 = warn(critical 措辞红线在 research-ethics 交付门);**措辞强度必须匹配证据强度**、 不显著只能报「未见显著差异」、

How do I install it?

Run `npx skills add Light0305/Light-skills --skill light-paper-writing --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.

Where does this skill come from?

From Light0305/Light-skills, a repository with 505 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.

Is a popular skill a good skill?

Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.

Keep going