wechat-article-production
Skill3 starsChanged 9 days agoMITView on GitHub
把 Markdown、粘贴正文或已有 handoff 制作成微信公众号文章,支持格式化、封面、正文配图、公众号排版、package 校验和创建草稿;也可将一组图片创建为微信贴图(newspic)草稿。按用户指定阶段续跑。配图默认原生调用 jiyi-little-dancer-illustrations,也可显式选择 baoyu-article-illustrator。
In this repository8 other files it gives its agents.
- agent-orchestratorSkill
- jiyi-little-dancer-illustrationsSkill
- lark-doc-copySkill
- session-analyzerSkill
- skill-doctorSkill
- skill-updaterSkill
- wx-video-downloaderSkill
- xhs-card-publisherSkill
The file
Files
微信公众号文章生产
目标
负责阶段调度、文件交接和最终门禁,不复制下游 Skill 的创作逻辑。
下游 Skill 必须通过当前运行时的原生 Skill 工具调用。读取下游 SKILL.md、运行它的内部脚本、让子 Agent 模仿它,均不算调用。没有原生 Skill 工具时,立即将运行标记为 blocked,不要手工降级。
阶段调度和调用细节见 references/stage-dispatch-rules.md。完整阶段契约见 references/execution-contract.md,配图引擎见 references/illustration-engines.md。首跑踩坑实录与修复手册见 references/known-pitfalls.md,任一阶段报错先查它再重试。主 Skill 只保留本文件的调度概览。
用户明确要“微信贴图 / 图片消息 / 小绿书草稿”时,不执行公众号长文链路;读取 references/image-text-draft.md,以 cards/ 图片目录和纯文本说明创建 newspic 草稿。该模式只创建草稿,禁止调用正式发布接口。
阶段范围
默认执行:
prepare → format → cover → illustrate → typeset → validate → publish按用户请求解析范围:
- “只做 X”:执行
prepare → X,不自动进入后续生产阶段; - “从 X 开始”:从
X执行到publish,除非用户明确说不发布或指定停止阶段; - “从 X 到 Y”:执行包含首尾的阶段;
- “跳过 X”:不执行 X,但不能假设它的产物已经存在;
- 阶段和 handoff 不明确时先询问,不猜测。
prepare 始终执行,用来创建或读取 package、解析范围、确定配图 Skill 和初始化 manifest;它不代表其他阶段已经完成。
配图选择
默认使用 jiyi-little-dancer-illustrations。用户明确请求或显式指定时,切换到 baoyu-article-illustrator。
选择优先级:显式 --illustration-skill > 用户当前请求 > 默认值。
详情见 references/illustration-engines.md。
Package 与 manifest
新运行统一创建到:
/Users/jiyi/Workspace/exports/wechat-articles/{slug}-{YYYYMMDD-HHmmss}/不要让主 Agent 自己拼接目录、复制源文件或初始化。使用一次性 prepare 脚本:
node <THIS_SKILL_ROOT>/scripts/prepare-package.mjs \
--slug <ARTICLE_SLUG> \
--source <SOURCE.md> \
--illustration-skill <ILLUSTRATION_SKILL> \
[--cover-text-override none|title-only|title-subtitle|text-rich]脚本成功时输出且仅输出绝对 package 路径。失败时保留脚本报错信息,不再尝试自行拼接命令或路径。
统一执行循环
对每个选中的阶段:
- 登记 stage-start
- 加载并执行下游 Skill
- 下游完成后运行机器校验
- 登记 stage-complete
各阶段的具体调用指令、命令和 prompt 模板见 stage-dispatch-rules.md。原生 Skill 调用规则、容错与恢复见 execution-contract.md。
Package 文件清单
package 中只保留以下稳定 handoff:
source.md
analysis.md
formatted.md
format-validation.json
cover/cover.png
cover/prompts/*
cover/prompt-validation.json
illustrations/outline.md
illustrations/outline-validation.json # 仅小舞伴
illustrations/prompts/*
illustrations/prompt-validation.json # 仅小舞伴
illustrations/image-validation.json # 仅小舞伴,机器图片门禁
illustrations/*.{png,jpg,jpeg,webp}
illustrations/ip-reference.png # 仅小舞伴
illustrations/illustration-handoff.json # 仅小舞伴
article-illustrated.md
article.html
article-preview.html
execution-manifest.json
validation-report.json
publish-result.json贴图模式不要求上述长文 package 文件;发布器会把图片、标题、说明和结果写入用户指定的产物目录。
完成报告
只报告 manifest 中真实存在的内容:
- 标准化后的阶段范围和明确跳过的阶段;
- 实际调用的配图 Skill,以及下游报告的图像后端;
- package 路径;
execution-manifest.json、validation-report.json和已生成产物路径;- 是否真的创建了公众号草稿。
没有原生调用记录、缺少 handoff、缺少实际文件或校验失败时,明确报告 blocked / failed,不要写“已完成”。
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

