laohan-fengmianqiuzhi skillA
laohan-fengmianqiuzhi is agent-read markdown (skill) from hanzhcn/laohan-skills: 根据真人AI口播稿和 Jeffrey 身份参考图，按文案冲突选择1个最匹配模板，生成1张9:16视觉母版，并产出3:4/4:3/16:9三张真实尺寸封面，共享到四平台7个逻辑入口。Use when 用户说“生成封面提示词”“做封面”“封面”“封面图”“封面词”或工作流进入⑥。.
Indexed from public GitHub and served as immutable, content-addressed versions. Install it pinned to an exact SHA-256 with the mdr CLI, and every file is verified against the hash recorded here before it reaches your agent. The deterministic audit below grades the latest version, and the same file always earns the same grade.
What the file says
# 真人口播封面直接生成 目标：从口播稿提炼一个跨平台可成立的点击钩子，只使用秋芝方向，按文案冲突选择1个最匹配模板，直接生成1张“人物＋场景＋准确中文标题”的9:16视觉母版，再以同一视觉主张按真实画布重新构图生成3张真实尺寸封面（3:4/4:3/16:9），共享到四平台7个逻辑入口。三个比例必须分别重新构图和生成，不把裁切或发布页临时取景当成适配。不要拆成无字底图、排字脚本、评分器或多阶段合成。 ## 输入与边界 - `episode-config.schema_version` 为4；身份reference或稿件绑定不满足时停止，不自行改合同。 - 读取当前 `01-口播稿.md`、`12-发布/多平台发布内容.md`和锁定的抖音发布信息。 - 人物参考固定为本期 `05-封面/reference/jeffrey-reference.jpg`，并作为 image reference 传给生图模型。 - ⑥先生成3张9:16完整候选，再为默认01生成3张共享真实尺寸封面；Jeffrey明确改选02或03时，⑫前必须为新rank重新生成这3张，旧rank成品不能混用。不允许简单裁掉人物、主物件或标题。 - 不在 prompt 中写“模仿某博主风格”。只使用 `references/douyin-cover-study.md` 的秋芝观察和 `references/qiuzhi-template-library.md`。 - 默认只生成3个秋芝方向候选；拉斐尔和柱子哥方向暂不进入默认输出。 - 三张按推荐程度从高到低排列为`01 → 02 → 03`。没有Jeffrey明确改选时，后续发布自动化优先使用`01`作为发布封面的源候选。 ## 执行 ### 1. 提炼封面文字 从口播稿提炼： - `主标题`：4—10个汉字，最多两行；保留真正影响理解的产品名。 - `辅助短句`：0—8个汉字；主标题已经说清楚时不要添加。 - `画面事件`：Jeffrey正在做什么，他面对什么具体问题或结果。 禁止“AI神器”“效率翻倍”等可套用到任何选题的空话。封面文字不能超出口播稿事实。 ### 2. 选择1个最匹配的秋芝模板 完整读取 `references/douyin-cover-study.md` 的秋芝部分和 `references/qiuzhi-template-library.md`，先按口播核心冲突筛选模板。 1. 先在内部筛出语义最匹配的5—8个模板。 2. 最终选择**1个**语义最命中本期钩子的模板；输出为单张封面（v1.19起不再产出3张排序候选）。 3. 遵守`1个基础模板 + 最多1个主物件机制`。 4. 上一期已用过的模板族原则上避免重复（防重复规则）。 …
Read the whole file at its exact version.
How to install
mdr add hanzhcn/laohan-skills/laohan-fengmianqiuzhi@v1.19.0mdr add hanzhcn/laohan-skills/laohan-fengmianqiuzhi@sha256:c677f33eb2b81e3bPin to a label to follow the author's releases, or to a sha256 to freeze the exact bytes forever. Either way the resolved hash is written to mdr.lock, and mdr install reproduces it on any machine.
[](https://markdownregistry.com/a/art_whv7cxzj4vmeffce)
1 badge views in 30 days
Versions
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| v1.19.0 latest | 2026-09-20 | 489f7c1 | 19,482 B | A | view · diff |
| v1.17.0 | 2026-09-20 | bba3237 | 15,087 B | A | view · diff |
| v1.17.0 | 2026-09-20 | bbab9be | 14,219 B | A | view · diff |
| v1.17.0 | 2026-09-20 | ef3f43e | 13,654 B | A | view · diff |
| v1.17.0 | 2026-09-20 | 9acdcf8 | 12,637 B | A | view |
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (19482 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No link to a raw IP address
- pass: No script tag
Source
hanzhcn/laohan-skills · 11 stars · license MIT · pushed 2026-09-20 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_whv7cxzj4vmeffce GET https://markdownregistry.com/api/v1/resolve?ref=hanzhcn/laohan-skills/laohan-fengmianqiuzhi GET https://markdownregistry.com/api/v1/blob/c677f33eb2b81e3bb72bcca0cd1ac6783a7c1e381bf8a9bbb2e3c0ff5026a015
Your agent does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.