← 返回 Skills 市场
ZeeLin 小红书自动发布
作者
kelcey2023
· GitHub ↗
· v1.0.99
· MIT-0
579
总下载
1
收藏
1
当前安装
6
版本数
在 OpenClaw 中安装
/install zeelin-xiaohongshu-autopost
功能描述
ZeeLin小红书自动化内容运营技能：从热文抓取→热梗生成→产品结合→发布文案的全流程自动化内容生产。 适用场景包括但不限于：短视频选题策划、爆款文案生成、钩子开头撰写、 账号人设定位、自媒体内容日历规划。当用户提到"内容运营"、"选题"、"爆款"、"钩子"、 "文案"、"短视频策划"、"自媒体运营"、"账号定位...
安全使用建议
This skill contains working scripts that: generate Xiaohongshu copy via the OpenClaw agent CLI and then automate the Xiaohongshu web UI to publish using a locally logged-in browser session. Before installing or running it: 1) Treat the scripts as requiring additional tools (openclaw/agent CLI, an agent-browser relay, jq, rg) and a logged-in creator.xiaohongshu.com browser — these are not declared in the metadata. 2) Audit and test in a safe sandbox account: do not run against your main account until you've verified behavior. 3) Require an explicit confirmation step before any publish action (remove or modify automatic triggering in SKILL.md and add a manual approval prompt). 4) Replace hardcoded paths (ROOT_DIR, default cover) with configurable parameters and declare any env vars needed. 5) Be aware the skill uses remote LLM calls (openclaw agent) and will send prompts/content to whatever model service that CLI is configured to use. 6) Consider platform/TOS and ethical issues: the generated content is designed to be persuasive/viral and automated posting increases risk of inappropriate or policy-violating posts. If you want to proceed, ask the skill author to (a) declare required binaries/env vars, (b) remove broad auto-trigger rules or require user approval, and (c) remove hardcoded local paths.
功能分析
Type: OpenClaw Skill
Name: zeelin-xiaohongshu-autopost
Version: 1.0.99
The skill bundle is a legitimate automation tool for generating and publishing Xiaohongshu (Red) social media content. It uses shell scripts (auto_topic_post_xiaohongshu.sh and post_xiaohongshu.sh) to orchestrate content generation via an LLM and automate browser interactions for posting. The logic is transparent, aligns with the stated marketing purpose, and contains no evidence of data exfiltration, malicious execution, or harmful prompt injection.
能力评估
Purpose & Capability
Skill claims no required binaries or env vars, yet the included scripts invoke external CLIs (openclaw agent, an agent-browser CLI via OPENCLAW_CLI), and use utilities like jq and rg. The scripts also hardcode a local ROOT_DIR (/Users/youke/...), default asset paths, and write drafts to disk; these expectations are not declared in the skill metadata and are disproportionate to the manifest's 'no requirements' claim.
Instruction Scope
SKILL.md instructs broad automatic triggering when users mention many keywords (even casual requests), and the scripts implement end-to-end generation + UI automation to publish on Xiaohongshu. That means the agent could autonomously generate and push posts using the user's browser session. The runtime instructions and scripts reference only the skill's docs and outputs, but they give the agent high discretion (web_search + LLM prompt) and perform actions that affect an external account; this is scope-expanding compared to a read/write-only drafting helper.
Install Mechanism
There is no install spec (instruction-only plus scripts), which is low-risk from an installation perspective. The code files are present but nothing in the package fetches external archives or runs an installer.
Credentials
The manifest lists no required environment variables, but scripts expect and use: the 'openclaw' CLI, an 'agent-browser' CLI (or $OPENCLAW_CLI), jq, rg (ripgrep), and read/write access to a specific local path and assets. They also rely on the user's browser session (cookies/auth) for publishing — effectively using the user's credentials implicitly. These requirements are not declared and grant the skill potential to act with the user's account privileges.
Persistence & Privilege
always:false (good). However, combined factors raise concern: SKILL.md encourages triggering on many user phrases, the scripts perform automated publishing via the user's browser session, and the agent has normal autonomous invocation enabled. This creates a realistic risk that the skill could generate and push content without an explicit, separate confirmation step.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install zeelin-xiaohongshu-autopost - 安装完成后，直接呼叫该 Skill 的名称或使用
/zeelin-xiaohongshu-autopost触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.99
## zeelin-xiaohongshu-autopost 1.0.99 更新日志
- 更新 `scripts/post_xiaohongshu.sh`，优化自动发布流程脚本。
- 稳定发布流程，仅接受标题和正文参数，更清晰地与主题-发布一体化脚本区分。
- 保持核心功能，未引入新参数或业务流程变动。
v2.0.0
热点抓取、热梗生成、产品结合、发布文案完整链路
v1.0.3
Fix typo: Zleelin -> ZeeLin; sync skill workflow metadata.
v1.0.2
Remove phrase: 浙江文旅账号运营 from skill introduction.
v1.0.1
Add Zleelin alias in metadata for improved keyword search discoverability.
v1.0.0
Initial release: xiaohongshu autopost flow (publish note -> long form -> new creation -> format -> next -> publish).
元数据
常见问题
ZeeLin 小红书自动发布 是什么？
ZeeLin小红书自动化内容运营技能：从热文抓取→热梗生成→产品结合→发布文案的全流程自动化内容生产。 适用场景包括但不限于：短视频选题策划、爆款文案生成、钩子开头撰写、 账号人设定位、自媒体内容日历规划。当用户提到"内容运营"、"选题"、"爆款"、"钩子"、 "文案"、"短视频策划"、"自媒体运营"、"账号定位... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 579 次。
如何安装 ZeeLin 小红书自动发布？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install zeelin-xiaohongshu-autopost」即可一键安装，无需额外配置。
ZeeLin 小红书自动发布 是免费的吗？
是的，ZeeLin 小红书自动发布 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
ZeeLin 小红书自动发布 支持哪些平台？
ZeeLin 小红书自动发布 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 ZeeLin 小红书自动发布？
由 kelcey2023（@kelcey2023）开发并维护，当前版本 v1.0.99。
推荐 Skills