Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add zhylq/yuan-skills --skill zhy-wechat-writinggit clone --depth 1 https://github.com/zhylq/yuan-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/zhylq/yuan-skills/zhy-wechat-writing)<a href="https://agentmods.dev/skills/zhylq/yuan-skills/zhy-wechat-writing"><img src="https://agentmods.dev/badge/skills/zhylq/yuan-skills/zhy-wechat-writing/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/zhylq/yuan-skills/zhy-wechat-writing"><img src="https://agentmods.dev/badge/skills/zhylq/yuan-skills/zhy-wechat-writing.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00036 | $0.06324 |
| Opus 5.5 | $0.00014 | $0.02530 |
| Sonnet 5.5 | $0.00007 | $0.01265 |
| Haiku 4.5 | $0.00004 | $0.00632 |
Grade A, and why
zhy-wechat-writing scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 582 lines — stays where its author put it; the contents beside it link to each section on GitHub.
微信公众号写作系统
Purpose
根据用户提供的主题(可选参考URL),自动完成公众号文章写作全流程:多来源检索与证据池整理、初稿生成、自审润色、参考资料整理,并可选自动配图与保存到公众号草稿箱(不提交发布)。
When to Use
- 用户请求"写一篇关于XXX的公众号文章"
- 用户请求"生成公众号文章,主题是XXX"
- 用户需要完整的公众号文章创作流程
- 用户希望"写完后自动配图"或"写完后发到公众号草稿箱"
Prerequisites
执行前需要确认:
- 用户已提供文章主题(
topic) - 如有参考文章 URL,可一并提供(
urls) - 若
topic为纯中文且未提供slug,建议补充英文/拼音kebab-case目录名;否则会使用 ASCII 降级方案
Workflow
按照以下步骤顺序执行(产物默认落盘到 articles/<slug>/...,便于复跑与追溯):
Phase 0: Preflight
目标:确定可稳定复用的目录与路径规范
操作:
- 计算
slug- 若用户提供
slug:直接使用(推荐:英文/拼音kebab-case) - 若
topic含拉丁字母/数字:对其做kebab-case - 否则降级:
wechat-article-YYYYMMDD
- 若用户提供
- 创建目录:
articles/<slug>/articles/<slug>/sources/
- 规范:Markdown图片引用必须使用相对路径与
/分隔符
Step 1: 素材搜集
目标:搜集与主题相关的素材,并整理为可追溯的证据池
操作:
- 若用户提供
urls:并行使用webfetch获取内容,提取要点,并记录URL与可获得的发布日期 - 若用户未提供
urls:并行使用WebSearch做多来源检索(建议覆盖:官方文档 / X(Twitter) / Reddit / 技术论坛 / 微信公众号 / 工程实践)- official/authority:官方文档、标准/规范、权威媒体解读
- community:X(Twitter)、Reddit、论坛/讨论
- practice:GitHub issues、工程博客、案例复盘
- 推荐并行 query 模板(按需组合,尽量加年份/时间范围以强调近期):
{topic} official documentation/{topic} release notes 2025 2026{topic} site:x.com/{topic} site:twitter.com{topic} site:reddit.com/{topic} site:reddit.com/r/<subreddit>{topic} site:github.com issues/{topic} site:github.com discussions{topic} site:stackoverflow.com/{topic} site:news.ycombinator.com{topic} site:mp.weixin.qq.com(公众号){topic} 实战 复盘 踩坑 2025 2026(中文工程实践)
- 合并去重,按可信度分级(high/medium/low),形成证据池,落盘:
articles/<slug>/sources/evidence.md
工具映射:
- 本流程中的“搜索”使用:
WebSearch - 本流程中的“抓取网页内容”使用:
webfetch
关于 WebSearch(实现说明):
- 优先使用运行环境自带的
WebSearch工具。 - 若当前环境没有可用的
WebSearch:用webfetch抓取公开搜索结果页(SERP),从结果中提取 URL 列表后再并行webfetch正文内容。
证据池条目格式(每条必须包含):
- title
- url
- published_at(可得则写)
- source_type(official/community/practice)
- key_takeaways(3-6条要点,尽量可直接改写成正文素材)
- confidence(high/medium/low)
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2mo ago First seen · 582 lines · 36 tokens per session scan A 145a0cc7ef25
zhy-wechat-writing is a skill published in the GitHub repository zhylq/yuan-skills (35 stars, last pushed 7mo ago), licensed MIT. It adds 36 tokens to every session and 6,324 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
zhy-wechat-writing compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| zhy-wechat-writingzhylq/yuan-skills | Skill | 35 | 7mo ago | MIT | A | 36 | Claude Code, Codex |
| tong-jincheng-perspective863401402/she-love-me | Skill | 999 | 17d ago | MIT | A | 160 | Claude Code, Codex |
| langbot-devlangbot-app/LangBot | Skill | 19k | yesterday | Apache-2.0 | A | 136 | Claude Code, Codex |
| edgeone skill scannerTencent/AI-Infra-Guard | Skill | 7.4k | 2d ago | Apache-2.0 | A | 148 | Claude Code, Codex |
| langbot-plugin-devlangbot-app/LangBot | Skill | 19k | yesterday | Apache-2.0 | A | 115 | Claude Code, Codex |
| langbot-space-opslangbot-app/LangBot | Skill | 19k | yesterday | Apache-2.0 | A | 101 | Claude Code, Codex |
| langbot-skills-maintenancelangbot-app/LangBot | Skill | 19k | yesterday | Apache-2.0 | A | 49 | Claude Code, Codex |