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 agentmods add skills/jackjin1997/redpress/xhs-postnpx skills add jackjin1997/redpress --skill xhs-postgit clone --depth 1 https://github.com/jackjin1997/redpressWrote 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/jackjin1997/redpress/xhs-post)<a href="https://agentmods.dev/skills/jackjin1997/redpress/xhs-post"><img src="https://agentmods.dev/badge/skills/jackjin1997/redpress/xhs-post.svg" alt="Measured on agentmods" 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 | $0.00067 | $0.01465 |
| Opus 5 | $0.00034 | $0.00732 |
| Sonnet 5 | $0.00013 | $0.00293 |
| Haiku 4.5 | $0.00007 | $0.00146 |
Grade A, and why
xhs-post scanned grade A with 1 finding 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 3d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- xiaohongshu-mcp 服务运行中（`curl -s http://localhost:18060/mcp` 可达； How it starts
The opening of the file, as written. The whole thing — 82 lines — stays where its author put it; the contents beside it link to each section on GitHub.
xhs-post：小红书图文发布流程
前置条件
发布有两条路径：自动（MCP 驱动浏览器发）和人工交付页（生成本地 HTML 手动发）。下面两条只对自动路径生效——选人工交付页不需要 MCP/登录，别提前 拿登录态打扰用户。到第 7 步用户选了自动，再检查：
- xiaohongshu-mcp 服务运行中（
curl -s http://localhost:18060/mcp可达； 没跑则提示：./vendor/xiaohongshu-mcp-darwin-arm64后台启动） - 用 MCP 工具
check_login_status确认已登录；未登录提示用户跑./vendor/xiaohongshu-login-darwin-arm64扫码
流程
- 建帖子目录：
lab/posts/YYYY-MM-DD-<slug>/（lab/ 是嵌套的私有 repo， 真实帖子不进公开仓库；lab/ 不存在时提示 clone jackjin1997/redpress-lab）， 把用户草稿存为draft.md - 问改写力度（必须，AskUserQuestion）——不要自作主张选力度：
- L0 原样：一字不动，只做卡片拆分和排版
- L1 微调：只修错别字、语病、标点，不动表达
- L2 润色：保留语义和关键原句（用户的金句不许改没），调整结构和 小红书体节奏 ← 推荐默认
- L3 重写：保留核心观点，大刀阔斧重组表达
- L4 发散：以草稿为种子，扩展新角度新内容
- 文案优化（按选定力度 + 下方 guidelines）：先读
lab/notes/里的 经验笔记（账号基线覆盖通用默认），然后给用户 3 个标题候选 + 优化后正文- tags + 卡片拆分方案，对话迭代到用户满意
- 写 post.json（契约见
src/types.ts，标题 ≤20 字、正文 ≤1000 字、 tags 不带 #、首卡必须 cover） - 渲染：
bun scripts/render.ts lab/posts/<dir>/，然后 Read 每张 PNG 检查 文字溢出/截断，有问题改 post.json 重渲染 - 确认门（必须）：向用户展示最终标题、正文、tags 和所有卡片图， 明确询问"确认发布？"——用户未明确同意前绝不调发布工具 / 不生成交付页
- 选发布方式（AskUserQuestion）——不要默认替用户选：
- A 自动发布：MCP 驱动浏览器发，省事，但有风控敞口
- B 人工交付页：生成本地 HTML，文字一键复制 + 图片下载，你手动贴进 小红书发，零风控敞口 ← 想稳走这个
- 7A 自动：先过本文档顶部「前置条件」（服务在跑 + 已登录），再调 MCP
工具
publish_content，参数 title、content（纯正文，不要拼接 # 标签）、tags（字符串数组，不带 # 号，MCP 自动处理）、images（cards/*.png 的绝对路径数组，cover 在第一位） - 7B 人工：
bun scripts/handoff.ts lab/posts/<dir>/生成cards/handoff.html，把file://绝对路径给用户，让其在浏览器打开、 复制文字下载图、手动发布；等用户回来说"发好了"再进收尾
- 收尾：把发布结果（自动=成功/失败+时间；人工=用户确认已手动发布+时间）
追加到该帖目录
draft.md末尾；迭代中的文风反馈沉淀到lab/notes/； 在 lab/ 内 commit + push
数据追踪（用户说"追踪一下数据"时）
bun lab/scripts/track.ts lab/posts/<dir>/
按标题搜索精确匹配 → 互动快照（赞/藏/评/转）append 到
lab/metrics/<slug>.jsonl。展示本次数据和与上次快照的增量；
明显的规律（某类标题涨得快等）提议写进 lab/notes/。
文案风格 guidelines
- 标题：≤20 字，有钩子——疑问句/反差/数字皆可，避免标题党到失真
- 正文结构：开头 1-2 句抛问题或亮观点 → 中间分点展开（每段 2-3 行， 段间空行）→ 结尾抛互动问题引导评论
- emoji：节制使用，段首或关键词后，全文 5-10 个，不要每句都加
- 语气：第一人称、聊天感、有真实观点和立场；观点讨论帖要"留口子"—— 自己的判断 + 承认不确定性，引导大家来辩
- tags：3-6 个，混合大流量词（如 AI、创业）和精准垂类词
- 卡片拆分：cover 放最有张力的一句话（≤12 字最佳）；每张 content 卡 一个主题，points 3-6 条、每条 ≤40 字
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.
- 3d ago First seen · 82 lines · 67 tokens per session scan A 8c7994441731
xhs-post is a skill published in the GitHub repository jackjin1997/redpress (2 stars, last pushed 1mo ago), licensed MIT. It adds 67 tokens to every session and 1,465 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
xhs-cli
Headless-browser-based CLI skill for Xiaohongshu (小红书, RedNote, XHS) to search notes, read posts, browse profiles, like, favorite, comment, and publish from the terminal.
xiaohongshu-skill
小红书 / Xiaohongshu / RedNote AI Agent Skill。用 Python Playwright 搜索和读取内容、管理登录会话、发布图文/视频/长文、评论、点赞和收藏；默认输出 JSON，任何写操作都必须先获得用户确认。用户提到 xiaohongshu、小红书、rednote、小红书搜索、发到小红书、小红书笔记分析、小红书运营或小红书自动化时触发。.
xiaohongshu-matrices-cli
Use xiaohongshu-matrices-cli for ALL Xiaohongshu (Little Red Book, 小红书) operations — searching notes, reading content, browsing users, liking, collecting, commenting, following, and posting. Invoke whenever the user requests any Xiaohongshu interaction.
cv-update-review
Run a confirmation-gated CV update workflow for Word CV documents: set up a reusable profile, scan configured local activity sources plus public evidence such as PubMed through NCBI E-utilities and configured website/RSS feeds, prepare a CV update packet, ask the user to approve completed/public/accepted items, create…
cite-them-all
Academic reference agent that identifies claims needing citations, searches PubMed/bioRxiv/medRxiv via MCP tools, and adds properly formatted references to Markdown manuscripts.
md-to-xhs-cards
Convert a Markdown file into Xiaohongshu (Little Red Book) image cards with deterministic layout, preserving markdown structure order and embedded local images. Use when users ask for direct MD-to-XHS card conversion, markdown poster cards, 图文卡片拆分, or preserving original markdown text/format in social image outputs.