ZJU-REAL/Easel / skills/builtin/skill-cross-platform-publish
skill-cross-platform-publish
Verified跨平台一键发布:一份内容适配并发布到多个平台(小红书/抖音/B站/公众号/快手/视频号/知乎/今日头条)。 按各平台格式约束(字数/比例/标签/内容类型)适配内容,再逐个委派对应平台发布 SKILL。 当用户说"一键发布""同时发到多个平台""多平台发布""一稿多发""全平台发""发到抖音+小红书+B站" "分发到各平台"时使用。适配由 LLM 做,路由与约束检查由 scripts/publish_dispatch.py。
跨平台一键发布
一稿多发:按各平台规范适配同一份内容,再逐平台委派发布 SKILL。走
scripts/publish_dispatch.py(约束检查 + 平台→publisher 路由 + 派发计划)。 内容适配由你(LLM)完成,实际发布委派各平台 publisher SKILL。
支持平台 → 发布 SKILL
| 平台 | publisher SKILL |
|---|---|
| xiaohongshu | skill-xhs-publisher ✅ |
| douyin | skill-douyin-upload ✅ |
| wechat(公众号) | skill-wechat-publisher ✅ |
| bilibili | skill-bilibili-upload ✅ |
| kuaishou | skill-kuaishou-upload ✅ |
| weixin-channels(视频号) | skill-channels-upload ✅ |
| zhihu | skill-zhihu-publisher ✅ |
| toutiao(今日头条) | skill-toutiao-publisher ✅ |
python <skill>/scripts/publish_dispatch.py platforms 看全部平台及约束。
输入
内容清单 manifest(JSON):
{
"content": {"title": "...", "body": "...", "tags": ["...", "..."], "media_type": "video"},
"platforms": ["xiaohongshu", "douyin", "bilibili"]
}
执行流程
脚本路径:skills/builtin/skill-cross-platform-publish/scripts/publish_dispatch.py。
# 1) 生成派发计划(校验各平台越限 + 路由到 publisher)
python skills/builtin/skill-cross-platform-publish/scripts/publish_dispatch.py plan --manifest content.json
- 按计划为每个平台适配内容(你来做):
- 依
constraints_note与warnings精简标题/正文、删减标签、调整话题格式。 - 依
recommend_aspect决定是否用 video-reframe 转比例(如竖版转横版发 B站)。 - 语气/话题按平台生态本地化(小红书亲和+emoji、B站梗、知乎专业、抖音短平快)。
- 依
- 发布前人设检查(有 Profile 时,见 AGENTS.md「发布前人设一致性检查」):
对适配后的各平台内容跑 skill-persona-check,评分喂
python skills/shared/scripts/persona_gate.py check --score 85—— 低于 80 分时告知分数、偏离点和修改建议,但不阻断发布;用户已明确要发布就继续执行。 - 逐平台发布:调用
dispatch[i].publisher对应的发布 SKILL 执行。 - 发布后:skill-publish-notify 推送结果;skill-short-link 生成带 UTM 追踪短链;
并把本次发布登记进 manifest 供归因层消费:
python skills/shared/scripts/manifest.py record --topic 露营攻略 \ --layer publish --skill skill-cross-platform-publish --profile 户外达人 \ --upstream outputs/露营攻略/article.md --summary "已发 小红书/抖音/B站"发布记录已由 publisher 脚本在
--exec成功时自动落「内容日历(_schedule.json)+ skill-publish-log」,无需手动补记(重复会双记)。仅当要补--persona-score/--persona-verdict等元数据时,再单独调 skill-publish-log 追加。
Profile 感知
- 有 Profile:默认目标平台取
platforms.md已开通的账号;各平台语气/话题贴合style.md; 合规底线遵守preferences.md。 - 无 Profile:询问目标平台,用平台通用规范适配。
规则
- 不要一稿原样多发——每个平台按其约束与生态适配(这是本 SKILL 的核心价值)。
- 发布前用
plan检查越限,逐条消除 warnings 再发。 - 视频跨横竖平台先用 video-reframe 转比例,别直接发导致黑边/裁切。
- 每个平台的实际发布委派其 publisher SKILL;未部署 publisher 的平台先提示用户。
- 批量按时间发布见 skill-publish-scheduler。
参考来源
一稿多发的关键不是"复制粘贴"而是"按平台重适配"。本 SKILL 用约束注册表把各平台的字数/比例/ 标签/类型规则固化,路由到对应 publisher,适配交给 LLM,发布交给平台 SKILL。
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |