Content Pipeline Skill
Safe content automation with human-in-the-loop approval. Draft → Review → Approve → Post.
Setup
CODEBLOCK0
For cryptographic approval signatures (password-protected):
CODEBLOCK1
This creates:
- -
drafts/ — work in progress (one post per file) - INLINECODE1 — human reviewed, awaiting your revision
- INLINECODE2 — you revised, ready for another look
- INLINECODE3 — human-approved, ready to post
- INLINECODE4 — archive after posting
- INLINECODE5 — review and customize before use
- INLINECODE6 — feedback thread logs (not posted)
Your Permissions
✅ Can do:
- - Write to INLINECODE7
- Read all content directories
- Revise drafts based on feedback
- Move revised files to INLINECODE8
- Run
content list to see pending content
❌ Cannot do:
- - Move files to
approved/ (only the human can approve) - Post content
- Set INLINECODE11
Creating Content
One post per file. Each suggestion or draft should be a single post, not a collection.
File naming: INLINECODE12
Use frontmatter:
CODEBLOCK2
Tell the human: "Draft ready for review: content review <filename>"
The Review Loop
CODEBLOCK3
- 1. You write draft to INLINECODE14
- Human runs
content review <file>:
-
With feedback → file moves to
reviewed/, you get notified
-
No feedback → human is asked "Approve?" → moves to
approved/
- 3. If feedback: you revise and move to INLINECODE18
- Human reviews from
revised/:
- More feedback → back to
reviewed/
- Approve → moves to
approved/
- 5. Posting happens manually via INLINECODE22
After Receiving Feedback
When you get review feedback:
- 1. Read the file from INLINECODE23
- Apply the feedback
- Move the file to INLINECODE24
- Confirm what you changed
- (Optional) Add a note: INLINECODE25
Platform Guidelines
LinkedIn
- - Professional but human
- Idiomatic language (Dutch for NL audiences, don't be stiff)
- 1-3 paragraphs ideal
- End with question or CTA
- 3-5 hashtags at end
X (Twitter)
- - 280 chars per tweet (unless paid account)
- Punchy, direct
- 1-2 hashtags max
- Use threads sparingly
- If Firefox auth fails, you can paste
auth_token and ct0 manually
Manual cookie steps:
1) Open x.com and log in
2) Open DevTools → Application/Storage → Cookies → https://x.com
3) Copy auth_token and INLINECODE29
Reddit (experimental)
- - Treat as experimental; API and subreddit rules can change
- Requires
subreddit: in frontmatter - Title comes from frontmatter
title: (or first line if missing) - Match each subreddit's rules and tone
Commands Reference
CODEBLOCK4
Security Model
The security model separates drafting (AI) from approval/posting (human):
- - ✅ Agent drafts content
- ✅ Agent revises based on feedback
- ❌ Agent cannot approve (human approves via
content review) - ❌ Agent cannot post
Posting is handled manually via CLI — never by the agent directly.
Platform-specific security
| Platform | Auth Storage | Encrypted? | Password Required? |
|---|
| LinkedIn | Browser profile | ✅ Yes | ✅ Yes |
| X/Twitter |
Firefox tokens | ✅ Yes | ✅ Yes |
Both platforms require password to post. Tokens are extracted from Firefox and encrypted locally.
内容管道技能
安全的内容自动化,含人工审核批准流程。草稿 → 审核 → 批准 → 发布。
设置
bash
npm install -g agent-content-pipeline
content init . # 创建文件夹 + 全局配置(在当前目录)
如需加密批准签名(密码保护):
bash
content init . --secure
这将创建:
- - drafts/ — 进行中的工作(每个文件一篇帖子)
- reviewed/ — 人工审核,等待你的修改
- revised/ — 你已修改,准备再次审阅
- approved/ — 人工批准,准备发布
- posted/ — 发布后的存档
- templates/ — 使用前审阅和自定义
- .content-pipeline/threads/ — 反馈线程日志(不发布)
你的权限
✅ 可以执行:
- - 写入 drafts/
- 读取所有内容目录
- 根据反馈修改草稿
- 将修改后的文件移至 revised/
- 运行 content list 查看待处理内容
❌ 不可执行:
- - 将文件移至 approved/(仅人工可批准)
- 发布内容
- 设置 status: approved
创建内容
每个文件一篇帖子。 每条建议或草稿应为单篇帖子,而非合集。
文件命名:YYYY-MM-DD-<平台>-<简短标题>.md
使用前置元数据:
yaml
platform: linkedin # linkedin | x | reddit(实验性)
title: 可选标题
status: draft
subreddit: programming # Reddit 必需
你的内容在此。
告知人工:草稿已准备好审核:content review <文件名>
审核循环
drafts/ → reviewed/ → revised/ → approved/ → posted/
↑ │
└──────────┘
更多反馈
- 1. 你将草稿写入 drafts/
- 人工运行 content review <文件>:
-
有反馈 → 文件移至 reviewed/,你收到通知
-
无反馈 → 人工被询问批准? → 移至 approved/
- 3. 如有反馈:你修改后移至 revised/
- 人工从 revised/ 审核:
- 更多反馈 → 回到 reviewed/
- 批准 → 移至 approved/
- 5. 通过 content post 手动发布
收到反馈后
当你收到审核反馈时:
- 1. 从 reviewed/ 读取文件
- 应用反馈
- 将文件移至 revised/
- 确认你修改的内容
- (可选)添加备注:content thread <文件> --from agent
平台指南
LinkedIn
- - 专业但人性化
- 地道的语言(面向荷兰受众用荷兰语,不要生硬)
- 1-3段为佳
- 以问题或行动号召结尾
- 末尾加3-5个话题标签
X(Twitter)
- - 每条推文280字符(除非付费账户)
- 简洁有力,直击要点
- 最多1-2个话题标签
- 谨慎使用线程
- 如果 Firefox 认证失败,可手动粘贴 auth_token 和 ct0
手动 Cookie 步骤:
1) 打开 x.com 并登录
2) 打开开发者工具 → 应用/存储 → Cookie → https://x.com
3) 复制 auth_token 和 ct0
Reddit(实验性)
- - 视为实验性功能;API 和子版块规则可能变更
- 前置元数据中需包含 subreddit:
- 标题来自前置元数据 title:(若无则取第一行)
- 遵循各子版块的规则和语气
命令参考
bash
content list # 显示草稿和已批准内容
content review <文件> # 审核:反馈或批准
content mv <目标> <文件> # 将文件移至 drafts/reviewed/revised/approved/posted
content edit <文件> # 在编辑器中打开($EDITOR 或 code)
content post <文件> # 发布(需确认)
content post <文件> --dry-run # 预览而不发布
content thread <文件> # 向反馈线程添加备注
安全模型
安全模型将草拟(AI)与批准/发布(人工)分离:
- - ✅ AI 草拟内容
- ✅ AI 根据反馈修改
- ❌ AI 不可批准(人工通过 content review 批准)
- ❌ AI 不可发布
发布通过 CLI 手动处理——绝不通过 AI 直接操作。
平台特定安全
| 平台 | 认证存储 | 加密? | 需要密码? |
|---|
| LinkedIn | 浏览器配置文件 | ✅ 是 | ✅ 是 |
| X/Twitter |
Firefox 令牌 | ✅ 是 | ✅ 是 |
两个平台均需密码才能发布。令牌从 Firefox 提取并在本地加密。