Journal Cover Letter Skill v3.2 (Recommended)

Recommended release. v3.2 writes Research, Review, and Bibliometrics cover letters while keeping manuscript facts authoritative. For Bibliometrics it joins research-output analysis with science mapping, requires specific journal relevance, and keeps submission-system-only choices out of the letter.

当前推荐版本。v3.2 覆盖 Research、Review 和 Bibliometrics；专家信只用于学习思路，事实仍以稿件为准。文献计量学路线会结合研究产出与知识图谱，具体解释期刊适配，并删除不需要写入投稿信的系统选项。

Choose one file: .skill for one Codex Skill, Plugin ZIP for the complete Codex package, or SKILL.md for another AI. GitHub adds two Source code links automatically; most users can ignore them.

SHA-256: skill 81a33d4bda80fc19c1234badd958a77d514957f8a6272d30802e1ed6728240c9; plugin 080a4ceb38a09bf3621222d77a830d1ca17c033d0f386ba79b61ed7d8636fd8e; SKILL.md ca4cea66b3e6de7d189f293a0c5a51731a07f25bb56ace9869e45df9a2b36120.

Journal Cover Letter Skill v3.1 (Legacy Stable)

Stable historical release. v3.1 keeps bibliometric letters specific by preserving a paper's own field map, frontier signal, or direction of change before explaining what those results mean.

稳定历史版本。v3.1 防止文献计量学投稿信过度抽象：先保留论文特有的领域分类、前沿主题或方向变化，再解释这些结果的意义。

Choose one file: .skill, Plugin ZIP, or portable SKILL.md. GitHub's automatic Source code links can be ignored.

Journal Cover Letter Skill v3.0 (Major Milestone · Legacy Stable)

Major milestone and stable historical release. v3.0 adds Bibliometrics as a third writing route because a field map needs a different argument from ordinary Research or Review. It also separates the journal's official article label from the letter's reasoning route.

重大里程碑和稳定历史版本。v3.0 革命性地加入 Bibliometrics 第三条路线：文献计量学研究呈现的是领域结构和演变，不能照搬普通 Research 或 Review 的写法；同时区分期刊正式文章标签和投稿信论证路线。

Choose one file: .skill, Plugin ZIP, or portable SKILL.md. GitHub's automatic Source code links can be ignored.

Journal Cover Letter Skill v2.3 (Legacy Stable)

Stable historical release. Review letters now use a clearer “old reading → synthesis finding → changed judgment” contrast and avoid repeated questions when the supplied evidence is already complete.

稳定历史版本。Review 路线更清楚地说明“过去怎样理解 → 综合后发现什么 → 接下来应怎样判断”，并减少资料完整时的重复提问。

Choose one file: .skill for one Codex Skill, Plugin ZIP for the complete Codex package, or SKILL.md for another AI. GitHub adds two Source code links automatically; most users can ignore them.

Journal Cover Letter Skill Trainer v0.2.0

Stable companion tool. The Trainer runs a blind AI draft, reveals a permitted expert letter only after the baseline is sealed, extracts reusable reasoning, revises the Skill, and tests every candidate in a fresh expert-free context.

独立稳定工具。Trainer 会先盲生成并封存 AI 版本，再读取已授权的专家投稿信，提炼可迁移思路、修改 Skill，并要求每轮候选版本在没有接触专家信的新对话中测试。

Choose one file: Trainer .skill, Trainer Plugin ZIP, or portable SKILL.md. Expert letters remain comparison benchmarks, not unquestionable facts.

Journal Cover Letter Skill v2.2 (Legacy)

Choose one file / 请选择一个文件：

• journal-cover-letter-skill-v2.2.skill — standalone Codex Skill / 独立 Skill
• journal-cover-letter-plugin-v2.2.zip — Codex Plugin / Codex 插件
• SKILL.md — portable workflow for other AI tools / 供其他 AI 阅读的单文件流程

GitHub also shows two automatic source-code downloads below; most users can ignore them.
GitHub 还会自动显示两个源码下载项，普通用户可以忽略。

Recommended version. v2.2 puts the scientific phenomenon and conclusion back at the center of Research letters. It translates methods into capabilities, combines results into one finding cluster, and removes details that do not change editorial judgment.｜推荐版本。 v2.2 让科学现象和核心发现重新成为 Research 投稿信的中心：把方法转化为可信能力，把结果合并成一个结论，并删除不会改变编辑判断的细节。

SHA-256:

• Skill: ac6a37e5e247f42a3e07a63873ad5a3515e3aa0690bc228ca4e9b701a715ec13
• Plugin: e95df19266be0282a94a7b75e31d2ae15268218eeff4e81993b1b61ababb0ecb
• SKILL.md: 10bdb590ab378da28aec619005172f84b4bd8c79b334bf32755fb3e9f26f31c6

Journal Cover Letter Skill v2.1 (Legacy)

Choose one file / 请选择一个文件：

• journal-cover-letter-skill-v2.1.skill — standalone Codex Skill / 独立 Skill
• journal-cover-letter-plugin-v2.1.zip — Codex Plugin / Codex 插件
• SKILL.md — portable workflow for other AI tools / 供其他 AI 阅读的单文件流程

GitHub also shows two automatic source-code downloads below; most users can ignore them.
GitHub 还会自动显示两个源码下载项，普通用户可以忽略。

Legacy version. Strengthened Reviews with field diagnosis, a clear interpretation, and measured promotion.｜加强 Review 的领域诊断、新解释和适度推广。

SHA-256:

• Skill: 34161d752173b507b903d5ac229fc838244b2ca4526d1e9d18528e3982444e71
• Plugin: 336c27a85a9075c84f7585eea1af0d3725cfea37766b7ed7079bff0652bc21c1
• SKILL.md: e76abd91e77cb27f210b0fa3d95771d58b5fc2d2fb68b3273880500dba8f6057

Journal Cover Letter Skill v2.0 (Legacy)

Choose one file / 请选择一个文件：

• journal-cover-letter-skill-v2.0.skill — standalone Codex Skill / 独立 Skill
• journal-cover-letter-plugin-v2.0.zip — Codex Plugin / Codex 插件
• SKILL.md — portable workflow for other AI tools / 供其他 AI 阅读的单文件流程

GitHub also shows two automatic source-code downloads below; most users can ignore them.
GitHub 还会自动显示两个源码下载项，普通用户可以忽略。

Legacy version. Separated Research discovery stories from Review synthesis logic.｜将 Research 的科学故事与 Review 的综合逻辑分开。

SHA-256:

• Skill: d67bf6eb96772d651471a359c2df3371e88b3f8a7095b2733d0c7c0efd6e2ccb
• Plugin: b3ea0d5a88e8e6e758c4890eb0c0a1d0b79b65ebdf963ff8593608bf4206f7a5
• SKILL.md: 64c63dda843b66369a15bfefa3a772858b086c36f5e43f98a61859ff9d2a2c28

Journal Cover Letter Skill v1.0 (Legacy)

Choose one file / 请选择一个文件：

• journal-cover-letter-skill-v1.0.skill — standalone Codex Skill / 独立 Skill
• journal-cover-letter-plugin-v1.0.zip — Codex Plugin / Codex 插件
• SKILL.md — portable workflow for other AI tools / 供其他 AI 阅读的单文件流程

GitHub also shows two automatic source-code downloads below; most users can ignore them.
GitHub 还会自动显示两个源码下载项，普通用户可以忽略。

Legacy version. Added fact checking, current journal research, previous-letter permissions, and final checks.｜加入事实核对、期刊查询、旧信权限和最终检查。

SHA-256:

• Skill: 27662d695d26f954e0ef9fe601aacd6b3dd1db18428a0d7ecc7a5d39321ed63f
• Plugin: e8cb11d9e490151ba2ff8d08368f5f157666cbdc096d22f564f8edab8bfa6f04
• SKILL.md: 1a6751f3b9a47336a4191be6fcb82975fe75095f77ef1cb9f95e030a89441a0f