Agent skill

Article To Wechat Cover

by dracohu2025-cloud in dracohu2025-cloud/draco-skills-collection

从飞书文档或本地 Markdown 提炼文章主题与风格,调用 OpenRouter 的 Nano Banana / Gemini Flash Image 生成 2.35:1 的微信公众号封面图,并可选上传为微信封面素材。

MITAuto-check passedProductivity & Automation

Install Article To Wechat Cover

skills CLI
$ npx skills add dracohu2025-cloud/draco-skills-collection --skill article-to-wechat-cover -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install dracohu2025-cloud/draco-skills-collection article-to-wechat-cover --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/dracohu2025-cloud/draco-skills-collection.git skills-src && mkdir -p .claude/skills && cp -r skills-src/article-to-wechat-cover .claude/skills/article-to-wechat-cover && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
article-to-wechat-cover
GitHub stars
227
Token cost
~1.5k tokens
SKILL.md length
314 words
Files
9 (incl. scripts, assets)
Skills in repo
27
Repo updated
First seen
Licence
MIT

At a glance

从飞书文档或本地 Markdown 提炼文章主题与风格,调用 OpenRouter 的 Nano Banana / Gemini Flash Image 生成 2.35:1 的微信公众号封面图,并可选上传为微信封面素材。

  • Works in 5 steps: 飞书文档 → 本地 Markdown → 主题来自文章,不靠人工硬写 prompt → …
  • Tasks that involve Messaging and chat bots
  • SKILL.md covers 目标, 新增封面方法:汤底 + 佐料 + 文章巧思, 输入来源 and 输出, plus 8 more sections
  • Runs Python scripts from its folder; calls python3; reaches xxx.feishu.cn and g1mu6da08l.feishu.cn; needs OPENROUTER_API_KEY and WECHAT_APP_SECRET

What it does

Article To Wechat Cover is an agent skill from dracohu2025-cloud/draco-skills-collection. 从飞书文档或本地 Markdown 提炼文章主题与风格,调用 OpenRouter 的 Nano Banana / Gemini Flash Image 生成 2.35:1 的微信公众号封面图,并可选上传为微信封面素材。

Its SKILL.md is about 1.5k tokens, which your agent loads only when the skill is triggered. The skill folder holds 11 other files, including scripts and assets (for example `README.md`, `scripts/article_to_wechat_cover.py` and `scripts/run.py`).

It sits in Productivity & Automation, covering Messaging and chat bots, Image generation and Model routing and gateways. It works with WeChat, OpenRouter and Google Gemini. The licence is MIT.

When your agent uses it

  • Tasks that involve Messaging and chat bots
  • Tasks that involve Image generation
  • Tasks that involve Model routing and gateways

Example prompts

  • “/article-to-wechat-cover”

Requirements

  • Python 3
  • A credential in OPENROUTER_API_KEY
  • A credential in WECHAT_APP_SECRET

Workflow steps

5 steps, taken from the step headings in SKILL.md.

  1. 飞书文档
  2. 本地 Markdown
  3. 主题来自文章,不靠人工硬写 prompt
  4. 公众号封面优先“像封面”,不是“像 PPT 截图”
  5. 优先横向头图构图

What it can do on your machine

Read from SKILL.md and the folder at commit 26e8975. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships 2 files in scripts/ (Python), which the agent can run.

    Shell commands in SKILL.md call:

    • python3

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    Hosts in commands or code, which the agent is likely to contact:

    • xxx.feishu.cn
    • g1mu6da08l.feishu.cn

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names these keys or tokens, usually read from environment variables:

    • OPENROUTER_API_KEY
    • WECHAT_APP_SECRET

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Article To Wechat Cover loads about 1.5k tokens when it runs. Until then it costs about 34 tokens; SKILL.md has 314 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~34
When it runs · the whole SKILL.md, loaded when a task matches
~1.5k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from dracohu2025-cloud/draco-skills-collection at commit 26e8975, republished under its MIT licence (© dracohu2025-cloud). 314 words, ~1,497 tokens.

Download SKILL.mdSave it as .claude/skills/article-to-wechat-cover/SKILL.md (or your agent's skills folder). This skill also uses 8 other files; get the full folder from GitHub.
name
article-to-wechat-cover
description
从飞书文档或本地 Markdown 提炼文章主题与风格,调用 OpenRouter 的 Nano Banana / Gemini Flash Image 生成 2.35:1 的微信公众号封面图,并可选上传为微信封面素材。
version
0.1.0
author
Hermes Agent
license
MIT

Article to WeChat Cover

当用户想要:

  • 为一篇飞书文档生成微信公众号封面图
  • 为一篇本地 Markdown 文章自动生成公众号头图
  • 让封面图主题跟随文章内容自动变化
  • 复用 Nano Banana / OpenRouter 图片生成链路做公众号封面
  • 在用户确认最终 prompt 后,把生成图片自动插入飞书文档最顶部作为 Hero 图

就用这个 skill。

目标

这个 skill 会做三件事:

  1. 读取文章内容(飞书文档 URL / token,或本地 Markdown)
  2. 用 OpenRouter 文本模型提炼:
    • 文章核心主题
    • 文章视觉风格
    • 适合公众号封面的构图与主视觉
  3. 先输出最终版生图 prompt 给用户确认
  4. 只有确认后,才用 OpenRouter 图片模型(Nano Banana / Gemini Flash Image)生成 2.35:1 微信公众号封面图

默认要求:

  • 宽高比固定 2.35:1
  • 画面体现文章核心主题
  • 风格由文章内容语气、领域、表达风格共同决定
  • 默认不渲染多余文字,避免把公众号封面做成低质海报

新增封面方法:汤底 + 佐料 + 文章巧思

这次吸收小小东 X 文章里的封面方法:封面生成不要每篇都从零发散,而是固定一套“汤底”,再从可复用“佐料库”里选一种风格,最后只根据标题/主题改一两个关键元素。

固定汤底

核心规则:

  • 先理解文章为什么值得被看见。
  • 把复杂信息收束成一个清晰、有力、耐看的推荐理由。
  • 手机端 3 秒可读。
  • 信息少而准,留白有呼吸感。
  • 主视觉来自文章/对象本身的气质,不做泛化 AI 科技图。
内置 10 组佐料

当前脚本内置:

  • 建筑图纸美学
  • 科技莫兰迪
  • 力量编辑感
  • 东方文人
  • 电影片头
  • 园艺手账
  • 治愈线条
  • 产品手册
  • 艺术展览
  • 综艺预告

文本模型会先从这 10 组中选 1 组,输出:

  • selected_recipe_id
  • selected_recipe_reason
  • recipe_adaptations:只允许 1-3 个文章专属改动
  • cover_message:封面 3 秒内要传达的一句话
Prompt 分层

生成链路现在分三层:

  1. analysis brief:理解文章。
  2. review_spec:给人审的结构化封面方案,包含汤底、佐料、巧思。
  3. execution_spec:真正喂给图片模型的瘦身 JSON,只保留会影响画面的字段。

这样能避免把流程说明、内部字段、解释文字一股脑塞给图片模型。少点废话,图更稳。

输入来源

支持两种:

1) 飞书文档
bash
python3 scripts/run.py from-feishu-doc \
  --doc 'https://xxx.feishu.cn/docx/xxx' \
  --final-prompt-output /tmp/wechat-cover-prompt.txt
2) 本地 Markdown
bash
python3 scripts/run.py from-markdown \
  --input ./article.md \
  --final-prompt-output /tmp/wechat-cover-prompt.txt

输出

成功后会输出:

  • 本地封面图路径(仅在已确认并真正生图后)
  • 主题分析 JSON 路径(可选)
  • 图片 JSON spec 路径(可选)
  • 最终版生图 prompt(可打印,也可写入文件)
  • 可选 feishu_url
  • 可选 thumb_media_id

必须先确认最终 prompt

这个 skill 现在默认采用两阶段流程:

阶段 1:先生成最终版生图 prompt,给用户确认

推荐:

bash
python3 scripts/run.py from-feishu-doc \
  --doc 'https://g1mu6da08l.feishu.cn/docx/xxxx' \
  --final-prompt-output /tmp/wechat-cover-prompt.txt

或:

bash
python3 scripts/run.py from-markdown \
  --input ./article.md \
  --final-prompt-output /tmp/wechat-cover-prompt.txt

这一步会:

  • 先分析文章主题与风格
  • 生成最终版生图 prompt
  • 在终端打印 prompt 预览
  • 可选写入 --final-prompt-output
  • 不会真的调用图片模型生图

如果命令输出里看到:

text
generation_skipped=pending_user_confirmation

说明流程正常停在“待确认”阶段。

阶段 2:确认后,再真正生图

只有在用户确认最终 prompt 没问题后,才重新执行并加上:

bash
--confirm-generate

例如:

bash
python3 scripts/run.py from-feishu-doc \
  --doc 'https://g1mu6da08l.feishu.cn/docx/xxxx' \
  --final-prompt-output /tmp/wechat-cover-prompt.txt \
  --output ./wechat-cover.jpg \
  --confirm-generate

经验规则:

  • 不要在第一次分析出 prompt 后立刻自动生图
  • 必须先把 final prompt 发给用户审阅
  • 用户确认后,才允许 --confirm-generate

补充经验:

  • 图片模型实际可能返回 image/jpeg,即使你手动把输出文件写成 .png
  • 因此脚本应按返回的 mime_type 自动修正输出后缀,避免出现“JPEG 内容却使用 PNG 扩展名”的混淆
  • 如果用户显式传入了一个不匹配的输出路径,CLI 最好打印一个 output_adjusted_from=... 提示,便于排查

统一入口命令

飞书文档 → 生成封面
bash
python3 scripts/run.py from-feishu-doc \
  --doc 'https://g1mu6da08l.feishu.cn/docx/xxxx' \
  --output ./wechat-cover.png
本地 Markdown → 生成封面
bash
python3 scripts/run.py from-markdown \
  --input ./article.md \
  --output ./wechat-cover.png

常用参数

  • --output:输出图片路径
  • --upload-feishu:生成后上传飞书云盘
  • --upload-wechat-cover:直接上传成微信封面素材,返回 thumb_media_id
  • --analysis-json:导出文章主题分析 JSON
  • --dump-json-spec:导出最终图片 JSON 规格
  • --final-prompt-output:导出最终版生图 prompt,便于发给用户确认
  • --confirm-generate:只有用户确认最终 prompt 后,才允许真正生图
  • --insert-into-feishu-doc-top:生成后把 Hero 图自动插到飞书文档正文最顶部
  • --replace-existing-top-image / --no-replace-existing-top-image:插入 Hero 图时是否替换已有顶部图片
  • --image-model:覆盖默认图片模型
  • --text-model:覆盖默认文本模型
  • --provider-order:OpenRouter provider 顺序,默认 Vertex AI
  • --visual-style-hint:人工补充风格偏好
  • --must-include:逗号分隔,指定必须出现的视觉元素
  • --must-avoid:逗号分隔,指定必须避免的视觉元素
  • --allow-text-overlay:允许封面内出现少量标题文字(默认关闭)

与飞书文档 → 微信草稿箱 skill 的关系

这个 skill 不直接替代 feishu-doc-to-wechat-draft,而是作为它的前置封面生成步骤。

推荐组合方式:

  1. 先用本 skill 生成封面图
  2. 若需要,直接 --upload-wechat-cover 拿到 thumb_media_id
  3. 再把这个 thumb_media_id 交给 feishu-doc-to-wechat-draft

例如:

bash
python3 scripts/run.py from-feishu-doc \
  --doc 'https://xxx.feishu.cn/docx/xxx' \
  --final-prompt-output /tmp/wechat-cover-prompt.txt

确认 prompt 没问题后,如果你想把生成图直接插回飞书文档最顶部:

bash
python3 scripts/run.py from-feishu-doc \
  --doc 'https://xxx.feishu.cn/docx/xxx' \
  --final-prompt-output /tmp/wechat-cover-prompt.txt \
  --confirm-generate \
  --insert-into-feishu-doc-top

如果还要同时上传成微信封面素材:

bash
python3 scripts/run.py from-feishu-doc \
  --doc 'https://xxx.feishu.cn/docx/xxx' \
  --final-prompt-output /tmp/wechat-cover-prompt.txt \
  --confirm-generate \
  --insert-into-feishu-doc-top \
  --upload-wechat-cover

然后把输出的 thumb_media_id 用于:

bash
python3 ~/.hermes/skills/productivity/feishu-doc-to-wechat-draft/scripts/run.py publish-feishu-doc-default \
  --doc 'https://xxx.feishu.cn/docx/xxx' \
  --thumb-media-id 'MEDIA_ID'

设计原则

1. 主题来自文章,不靠人工硬写 prompt

脚本会先从标题、摘要、正文提炼主题与风格,再生成图片 prompt。

2. 公众号封面优先“像封面”,不是“像 PPT 截图”

默认避免:

  • 杂乱文字
  • UI 截图感
  • 信息图堆砌
  • 水印
  • logo 拼贴
3. 优先横向头图构图

固定 2.35:1,适合公众号封面横幅视觉。

依赖

  • OPENROUTER_API_KEY
  • 可选:OPENROUTER_TEXT(默认 google/gemini-3.1-flash-lite-preview)
  • 可选:OPENROUTER_IMAGE
  • 如需读取飞书文档:已可用的 lark-cli
  • 如需上传微信封面素材:
    • WECHAT_APP_ID
    • WECHAT_APP_SECRET

验证建议

先本地跑:

bash
python3 scripts/run.py from-markdown --input ./article.md --analysis-json /tmp/analysis.json --dump-json-spec /tmp/spec.json

确认:

  • analysis 是否抓住主题
  • spec 是否保持 2.35:1
  • 最终图片是否贴合文章主题

适用边界

适合:

  • 技术文章
  • 观点评论
  • 方法论总结
  • 产品/AI/效率类公众号文章

不适合直接拿来做:

  • 强品牌商业广告 KV
  • 多页海报
  • 需要精确排版文字设计的活动主视觉

这类任务应额外提供人工 art direction。

© dracohu2025-cloud, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 8 other files (scripts, assets) in article-to-wechat-cover of dracohu2025-cloud/draco-skills-collection.

  • SKILL.md
  • .env.example
  • .gitignore
  • README.md
  • assets/example-wechat-cover.jpg
  • requirements.txt
  • scripts/article_to_wechat_cover.py
  • scripts/run.py
  • tests/test_article_to_wechat_cover.py

Open the folder on GitHubat commit 26e8975

Compare with similar skills

Article To Wechat Cover next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Article To Wechat Cover compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Article To Wechat Cover this skilldracohu2025-cloud/draco-skills-collection227—~1.5kAutomated safety check: PassMIT
Md2wechatgeekjourneyx/md2wechat-skill3.7k—~3.8kAutomated safety check: PassCustom licence
AI Image Creatorevolution-foundation/evo-nexus545—~5.1kAutomated safety check: NotesCustom licence
Openrouter Imgrtadewald/skills185—~800Automated safety check: NotesNone
Image Genwpsnote/wpsnote-skills179—~1.3kAutomated safety check: PassNone
Nano Banana Pro Openroutergithub/awesome-copilot40k1 repos~664Automated safety check: PassMIT

Similar skills

  • Md2wechat

    geekjourneyx/md2wechat-skill

    Convert Markdown to WeChat Official Account HTML. An agent skill from geekjourneyx/md2wechat-skill.

    3.7k GitHub stars~3.8k tokensUpdated 17 days ago
    Media & CreativeAuto-check passed
  • AI Image Creator

    evolution-foundation/evo-nexus

    Generates PNG images through OpenRouter models, with transparent backgrounds and reference-image edits, and describes existing images with multimodal vision.

    545 GitHub stars~5.1k tokensUpdated 5 mo ago
    Media & CreativeAuto-check: notes
  • Openrouter Img

    rtadewald/skills

    Generate or edit images via OpenRouter Image API across top models (Nano Banana, GPT Image, FLUX.2, Seedream, Qwen, Grok).

    185 GitHub stars~800 tokensUpdated 12 days ago
    Media & CreativeAuto-check: notes
  • Image Gen

    wpsnote/wpsnote-skills

    AI 图像生成助手,支持文生图和图生图,对接 OpenRouter / 阿里云百炼 / 火山方舟 / Google Gemini。

    179 GitHub stars~1.3k tokensUpdated 4 mo ago
    Media & CreativeAuto-check passed
  • Nano Banana Pro Openrouter

    github/awesome-copilot

    Official

    Generate or edit images via OpenRouter with the Gemini 3 Pro Image model.

    40k GitHub starsUsed in 1 repo~664 tokens
    Media & CreativeAuto-check passed
  • Generate Image

    K-Dense-AI/claude-scientific-writer

    Generate or edit images with AI models through the OpenRouter Image API (Gemini, Seedream, Recraft, GPT-Image, Riverflow).

    2.4k GitHub starsUsed in 1 repo~3.8k tokens
    Media & CreativeAuto-check: notes

More from dracohu2025-cloud/draco-skills-collection

All 27 skills in this repo
  • AI News Bitable Archive Dingtalk

    dracohu2025-cloud/draco-skills-collection

    A skill your agent uses when parsing a DingTalk daily news document and syncing title, date, Top 3, summary, conclusion, document link, and status into a DingTalk multidimensional table.

    227 GitHub stars~795 tokensUpdated 23 days ago
    Auto-check passed
  • Feishu Seedance Video Pipeline

    dracohu2025-cloud/draco-skills-collection

    A skill your agent uses when running a Feishu/Lark Base-centered Seedance video production pipeline: inspect/backfill the Base row, create/reference assets, build Row-24-style Chinese Seedance…

    227 GitHub stars~5k tokensUpdated 23 days ago
    Auto-check: notes
  • Motion Canvas

    dracohu2025-cloud/draco-skills-collection

    Production pipeline for Motion Canvas — TypeScript-based programmatic vector animation with real-time preview.

    227 GitHub stars~1.5k tokensUpdated 23 days ago
    Auto-check passed
  • News Aggregator Skill

    dracohu2025-cloud/draco-skills-collection

    A skill your agent uses when collecting public news candidates from Hacker News, GitHub Trending, Hugging Face papers, and other web sources for briefings and daily reports.

    227 GitHub stars~596 tokensUpdated 23 days ago
    Auto-check passed
  • Open Design To Open Slide

    dracohu2025-cloud/draco-skills-collection

    A skill your agent uses when mining Open Design visual systems and converting them into Open Slide React template kits or standalone 20-page template albums, with build, deployment, screenshot QA…

    227 GitHub stars~2.9k tokensUpdated 23 days ago
    Auto-check passed
  • Daily AI Agent Aigc Top News Dingtalk

    dracohu2025-cloud/draco-skills-collection

    A skill your agent uses when generating a daily 24h AI / Agent / AIGC top-news briefing, publishing it as a DingTalk document, validating it, and optionally archiving it to a DingTalk…

    227 GitHub stars~3k tokensUpdated 23 days ago
    Auto-check passed

Questions about Article To Wechat Cover

What does Article To Wechat Cover do?

从飞书文档或本地 Markdown 提炼文章主题与风格,调用 OpenRouter 的 Nano Banana / Gemini Flash Image 生成 2.35:1 的微信公众号封面图,并可选上传为微信封面素材。. Article To Wechat Cover is an agent skill from dracohu2025-cloud/draco-skills-collection.

When should I use Article To Wechat Cover?

Article To Wechat Cover fits situations like: tasks that involve Messaging and chat bots; tasks that involve Image generation; tasks that involve Model routing and gateways.

How do I install Article To Wechat Cover in Claude Code?

Run `npx skills add dracohu2025-cloud/draco-skills-collection --skill article-to-wechat-cover -a claude-code`. Or copy the skill folder (article-to-wechat-cover in dracohu2025-cloud/draco-skills-collection) into .claude/skills/article-to-wechat-cover in your project. Claude Code loads it when a task matches its description.

How do I install Article To Wechat Cover in Codex?

Run `npx skills add dracohu2025-cloud/draco-skills-collection --skill article-to-wechat-cover -a codex`. Or copy the skill folder (article-to-wechat-cover in dracohu2025-cloud/draco-skills-collection) into .agents/skills/article-to-wechat-cover in your project. Codex loads it when a task matches its description.

Can I use Article To Wechat Cover in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add dracohu2025-cloud/draco-skills-collection --skill article-to-wechat-cover -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/article-to-wechat-cover, .gemini/skills/article-to-wechat-cover, .github/skills/article-to-wechat-cover and .opencode/skills/article-to-wechat-cover in your project.

What does Article To Wechat Cover need to run?

Going by SKILL.md and its folder, Article To Wechat Cover needs Python for the scripts in its folder, the command-line tools its instructions call (python3) and credentials named OPENROUTER_API_KEY and WECHAT_APP_SECRET. Our summary lists: Python 3; A credential in OPENROUTER_API_KEY; A credential in WECHAT_APP_SECRET.

Does Article To Wechat Cover access the network?

SKILL.md names 2 domains. In commands or code: xxx.feishu.cn and g1mu6da08l.feishu.cn; the agent is likely to contact these when it follows the instructions. This is read from the text; nothing was executed.

Is Article To Wechat Cover safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does Article To Wechat Cover use?

Article To Wechat Cover is published under the MIT licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Article To Wechat Cover use?

About 1.5k tokens (SKILL.md is roughly 6k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Article To Wechat Cover?

Skills that share tags, products or a category with Article To Wechat Cover: Md2wechat (geekjourneyx/md2wechat-skill, 3.7k stars), AI Image Creator (evolution-foundation/evo-nexus, 545 stars), Openrouter Img (rtadewald/skills, 185 stars) and Image Gen (wpsnote/wpsnote-skills, 179 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Article To Wechat Cover?

dracohu2025-cloud (a GitHub user) maintains it in dracohu2025-cloud/draco-skills-collection, which has 227 GitHub stars. The repository holds 27 skills in this directory. The repository was last updated on September 17, 2026.

Source: dracohu2025-cloud/draco-skills-collection on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.