Content Pipeline Skill

Safe content automation with human-in-the-loop approval. Draft → Review → Revise (via pi) → Approve → Post.

Setup

npm install -g agent-content-pipeline
content init .

This creates:

• drafts/ — write your drafts here
• reviewed/ — human reviewed, pi is rewriting
• revised/ — rewritten, ready for another look
• approved/ — human-approved, ready to post
• posted/ — archive after posting
• templates/ — platform templates

Your Permissions

✅ Can do:

• Write to drafts/
• Read all content directories
• Add notes: content thread <file> --from agent
• Run content list to see pending content

❌ Cannot do:

• Move files between folders (the pipeline does this automatically)
• Move files to approved/ or posted/ (human only)
• Post content
• Set status: approved

How the Pipeline Works

When the human reviews a draft and gives feedback, pi automatically rewrites the draft and moves it to revised/. You do not need to revise drafts yourself.

drafts/ → reviewed/ → revised/ → approved/ → posted/
  you       human      pi          human       human
  write     reviews    rewrites    approves    posts

Your job is to write good initial drafts. The rewrite loop is handled automatically.

Creating a Draft

File naming: YYYY-MM-DD-<platform>-<slug>.md

---
platform: linkedin    # linkedin | x | reddit | devto | hashnode
title: "Required for reddit, devto, hashnode"
status: draft
subreddit: programming  # Required for Reddit
tags: [tag1, tag2]      # Optional, used by devto/hashnode
---

Your content here.

Platform Guidelines

LinkedIn

• Professional but human
• 1-3 paragraphs ideal
• End with question or CTA
• 3-5 hashtags at end

X (Twitter)

• 280 chars per tweet
• Use --- to separate tweets in a thread
• Punchy, direct
• 1-2 hashtags max

Reddit (experimental)

• Title required in frontmatter
• Match each subreddit's rules and tone

dev.to / Hashnode

• Full markdown article
• Title required in frontmatter
• Tags as array of slugs

Commands Reference

content list                    # Show all folders with timestamps
content review <file>           # Review: give feedback (pi rewrites) or approve
content edit <file>             # Open in editor
content post <file>             # Post (prompts for confirmation)
content post <file> --dry-run   # Preview without posting
content thread <file>           # Add a note to the feedback thread
content platforms               # List available platforms

Security Model

• ✅ Agent drafts content
• ❌ Agent cannot approve or post (human only)

Posting is handled manually via CLI.