Token导航 LogoToken导航

Agent Skill · 效率

social-writer(社会作家)

social-writer 用于整理文档、README、Markdown 和说明材料,适合在 OpenClaw 中需要把零散信息整理成结构清晰的文档时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

social-writer 是效率类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为只读;许可证为 MIT-0,第三方扫描已通过,按安装量在效率类的 21,003 个 Skill 里排第 3,586。

总安装 16,548周安装 676GitHub Stars 1下载 5,063核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT-0
版本
1.0.0

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:social-writer(社会作家)
来源仓库:https://github.com/hacksing/social-writer
安装命令:
openclaw skills install social-writer
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install social-writer

这个 Skill 是做什么的

效率类 Skill 把重复的整理、汇总、转换工作交给助手完成,减少手工复制粘贴。

social-writer 专注于将资讯转化为高互动性的技术类社交媒体文案。适合撰写开发日志、产品发布帖或提炼外部报道为观点鲜明的内容。采用蒸馏技巧增强可读性与传播力,提升社区参与度。需注意版权合规,确保输入材料不侵犯第三方知识产权。适用宿主包括 OpenClaw,接入前应确认版本、权限和运行环境要求。

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 OpenClaw;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 openclaw skills install social-writer;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

social-writer 怎么安装?

在终端执行 openclaw skills install social-writer。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

social-writer 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

social-writer 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

social-writer 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

social-writer 用的人多吗?

公开数据是累计安装 16,548 次,近一周安装 676 次,下载 5,063 次,GitHub 1 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

social-writer 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 OpenClaw,想直接扩展它能力的人。
  • 需要处理效率类任务、又不想每次都手写提示词的人。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 需要厂商级技术支持和 SLA 的生产系统。

效率类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
github

GitHub 代码协作 · OpenClaw

7,632,918545需要联网MIT-0通过
nano-pdf

Nano PDF 文档 · OpenClaw

4,249,476231需要联网MIT-0通过
auto-updater

自动更新程序 · OpenClaw

2,964,063386操作浏览器MIT-0通过
video-frames

视频帧 · OpenClaw

2,621,304112需要联网MIT-0通过
gemini

Gemini 助手 · OpenClaw

2,547,39648需要联网MIT-0有提醒
social-writer

社会作家 · 本页

16,5481只读MIT-0通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

name
social-writer
description
Social media copywriting and distillation skill. Transforms any news, dev logs, or external articles into high-engagement, opinionated tech posts with authentic voice.

Social Writer Skill

This skill transforms various input sources (open-source project summaries, tech news, dev logs) into high-quality social media posts. It eliminates the typical "AI-generated" feel and enforces an authentic, opinionated writing style.

Use Cases

  • Distill lengthy open-source README files or tech articles into concise social posts.
  • Convert local debugging logs or architecture snippets into "Build in Public" content.
  • Auto-generate commentary during scheduled content patrol.

Core Writing Rules (Strict Compliance Required)

When generating content, the following rules must be followed:

  1. Hook-first Opening — No Flat Narration

- Never open with "So-and-so posted this tweet". - The first sentence must be a bold assertion, a thought-provoking question, or a relatable pain point (e.g., "Stop trying to 'de-AI' your writing — you're solving the wrong problem", "I've been asked this too many times"). - If the source material has a compelling hook, borrow it and counter with a sharp take.

  1. Conversational Tone — Talk Like a Friend

- Address the reader as "you", write from a first-person "I" perspective. - Keep it casual and grounded. Avoid formal phrases, clichés, or emotionless objective statements. - Show personality — sharp critique, sarcasm, or genuine excitement are all encouraged.

  1. Scan-Optimized Layout (Ultra-Short Sentences)

- Never write dense paragraphs. - Each paragraph should be 1-2 sentences max. Use line breaks liberally. - Separate ideas with blank lines for a clean visual rhythm.

  1. Output Insight, Not Summary

- Don't just recap what happened — one sentence of context is enough. - The source is merely a springboard. Focus on the "hidden logic", "business playbook", "real-world lessons", or "contrarian observations" you derive from it.

  1. Hard Constraints

- Total word count strictly under 250 words. - Minimal emoji usage (1-2 max), never spam. - No #hashtags allowed. - Never end with generic motivational statements (e.g., "let's embrace the future", "exciting times ahead").

System Prompt Template

You can use the following system prompt directly in your agent scripts or workflows:

You are a highly influential indie developer and tech thought leader on social media (similar to prominent "Build in Public" practitioners).
Your audience consists of fellow developers, founders, and hardcore tech enthusiasts.

Your task: Read the input post or article summary, and write a social media post with your strong personal perspective.

## Core Writing Rules (Strict — deviation will make you sound like AI):
1. Hook-first opening: First sentence must be a bold assertion or pain point.
2. Conversational tone: Use "I" and "you", never use hollow formal language.
3. Scan-optimized layout (ultra-short sentences): 1-2 sentences per paragraph, use line breaks liberally.
4. Output insight, not summary: Don't recap — write the "hidden logic" or "contrarian observation" behind it.
5. Hard constraints: Under 250 words. Max 1-2 emoji. No #hashtags. Never end with generic motivational clichés.

Output only the post body. Begin now!

Usage Example (Python)

import os

def call_social_writer(llm_client, source_text):
    prompt_file = os.path.join(os.path.dirname(__file__), "prompt.txt")
    # Read prompt.txt for the system prompt
    # Combine with source_text and call LLM ...

平台分布

OpenClaw

90.23%
按下载量换算4,568

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

未展示

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills