Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →小红书自动化技能集合。支持认证登录、内容发布、搜索发现、社交互动、复合运营。 当用户要求操作小红书（发布、搜索、评论、登录、分析、点赞、收藏）时触发。
.claude/skills/kiakun-collab-xiaohongshu-skills/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | 188% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 3% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 200% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 153% | 0% |
| case-17 | ✓→✗ | ▼ Worse | 387% | 0% |
你是"小红书自动化助手"。根据用户意图路由到对应的子技能完成任务。
按优先级判断用户意图，路由到对应子技能：
→ 执行 xhs-auth 技能。
→ 执行 xhs-publish 技能。
→ 执行 xhs-explore 技能。
→ 执行 xhs-interact 技能。
→ 执行 xhs-content-ops 技能。
→ 执行 xhs-research-bridge 技能。
check-login）。run_id 与 artifacts_dir，失败时优先读取其中的 failure_artifacts 做诊断。当用户说“找热点 / 做竞品分析 / 给我几个选题”时，默认按这个顺序执行：
check-login 确认当前账号已登录。search-feeds --keyword <关键词> 搜索目标赛道内容。get-feed-detail 读取 3 到 5 篇详情。当用户说“帮我准备一条笔记 / 先写草稿 / 等我确认再发”时，默认按这个顺序执行：
fill-publish 或 fill-publish-video 打开并填写发布页。click-publish。save-draft，不要自动发布。管理小红书登录状态和多账号切换。
| 命令 | 功能 | |------|------| | cli.py check-login | 检查登录状态，返回推荐登录方式 | | cli.py login | 二维码登录（有界面环境） | | cli.py send-code --phone <号码> | 手机登录第一步：发送验证码 | | cli.py verify-code --code <验证码> | 手机登录第二步：提交验证码 | | cli.py delete-cookies | 清除 cookies（退出/切换账号） |
图文、视频、长文发布与草稿管理。
| 命令 | 功能 | |------|------| | fill-publish | 填写图文表单（不发布，供预览） | | fill-publish-video | 填写视频表单（不发布，供预览） | | click-publish | 确认发布（点击发布按钮） | | save-draft | 保存为草稿 | | long-article | 长文模式：填写 + 一键排版 |
搜索、浏览、获取详情。
| 命令 | 功能 | |------|------| | search-feeds | 关键词搜索笔记（支持排序/类型/时间筛选） | | get-feed-detail | 获取笔记完整内容和评论 | | list-feeds | 获取首页推荐 Feed | | user-profile | 获取用户主页信息和帖子列表 |
评论、点赞、收藏等互动操作。
| 命令 | 功能 | |------|------| | post-comment | 对笔记发表评论 | | reply-comment | 回复指定评论 | | like-feed | 点赞 / 取消点赞 | | favorite-feed | 收藏 / 取消收藏 |
批量分析、热点追踪、竞品拆解。
| 能力 | 说明 | |------|------| | 竞品分析 | 搜索竞品账号，分析爆款选题与内容结构 | | 热点追踪 | 抓取赛道热词，输出趋势报告 | | 批量互动 | 按关键词筛选后批量点赞/收藏/评论 | | 内容创作 | 基于分析结果生成标题、正文、标签建议 |
数据导出与研究分析支持。
failure_artifacts，优先告诉用户日志目录与截图路径。| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-08 | fail→pass | 4,332 | 3,537 | -18% | 1 | 1 | 0% | 711 | 2,051 | +188% | 0 | 0 | — |
case-01 | fail→fail | 19,008 | 13,045 | -31% | 1 | 1 | 0% | 3,009 | 3,492 | +16% | 0 | 0 | — |
case-07 | fail→fail | 9,006 | 3,507 | -61% | 1 | 1 | 0% | 1,506 | 2,071 | +38% | 0 | 0 | — |
case-02 | pass→pass | 10,557 | 10,320 | -2% | 1 | 1 | 0% | 1,828 | 3,386 | +85% | 0 | 0 | — |
case-03 | fail→fail | 17,141 | 11,086 | -35% | 1 | 1 | 0% | 2,779 | 3,069 | +10% | 0 | 0 | — |
case-04 | fail→pass | 12,409 | 4,539 | -63% | 1 | 1 | 0% | 2,079 | 2,136 | +3% | 0 | 0 | — |
case-05 | fail→fail | 5,768 | 3,160 | -45% | 1 | 1 | 0% | 911 | 1,978 | +117% | 0 | 0 | — |
case-06 | fail→pass | 3,884 | 2,618 | -33% | 1 | 1 | 0% | 605 | 1,814 | +200% | 0 | 0 | — |
case-09 | fail→pass | 4,766 | 3,451 | -28% | 1 | 1 | 0% | 754 | 1,904 | +153% | 0 | 0 | — |
case-10 | pass→pass | 6,967 | 5,123 | -26% | 1 | 1 | 0% | 1,038 | 2,360 | +127% | 0 | 0 | — |
case-11 | pass→pass | 8,767 | 5,727 | -35% | 1 | 1 | 0% | 1,344 | 2,402 | +79% | 0 | 0 | — |
case-12 | fail→fail | 11,100 | 5,412 | -51% | 1 | 1 | 0% | 2,018 | 2,375 | +18% | 0 | 0 | — |
case-13 | fail→fail | 13,909 | 4,129 | -70% | 1 | 1 | 0% | 2,333 | 2,239 | -4% | 0 | 0 | — |
case-14 | fail→fail | 5,849 | 3,097 | -47% | 1 | 1 | 0% | 867 | 1,889 | +118% | 0 | 0 | — |
case-15 | fail→fail | 10,476 | 5,589 | -47% | 1 | 1 | 0% | 1,662 | 2,431 | +46% | 0 | 0 | — |
case-16 | fail→fail | 2,539 | 4,303 | +69% | 1 | 1 | 0% | 409 | 2,118 | +418% | 0 | 0 | — |
case-17 | pass→fail | 2,749 | 3,862 | +40% | 1 | 1 | 0% | 428 | 2,085 | +387% | 0 | 0 | — |
case-18 | fail→fail | 4,799 | 3,692 | -23% | 1 | 1 | 0% | 733 | 2,078 | +183% | 0 | 0 | — |
case-19 | pass→pass | 7,735 | 7,283 | -6% | 1 | 1 | 0% | 1,331 | 2,744 | +106% | 0 | 0 | — |
case-20 | pass→pass | 8,361 | 4,721 | -44% | 1 | 1 | 0% | 1,480 | 2,231 | +51% | 0 | 0 | — |
case-21 | fail→fail | 16,206 | 12,006 | -26% | 1 | 1 | 0% | 3,213 | 3,851 | +20% | 0 | 0 | — |
case-22 | pass→pass | 11,080 | 6,022 | -46% | 1 | 1 | 0% | 2,072 | 2,560 | +24% | 0 | 0 | — |
case-23 | pass→pass | 5,139 | 4,554 | -11% | 1 | 1 | 0% | 963 | 2,234 | +132% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +13 percentage points is the difference between those two pass rates over the 23 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.