openclaw-content-kit

Safe content automation for AI agents. Draft → Review → Revise → Approve → Post.

The pattern: Your AI drafts content. You review and give feedback. They revise. You approve. You post.

Built for OpenClaw — your agent gets notified automatically when you give feedback.

Why?

AI agents shouldn't post directly to social media. Too risky. But they're great at drafting.

This kit enforces human-in-the-loop:

• Agent → writes drafts, revises based on feedback
• Human → reviews, gives feedback, approves, posts

Install

npm install -g openclaw-content-kit

Includes built-in posters for LinkedIn and X/Twitter.

Quick Start

# 1. Initialize in your workspace (creates folders + global config)
content-kit init

# 2. Authenticate (once per platform)
content-kit auth linkedin    # Opens browser for login
content-kit auth x           # Extracts tokens from Firefox

# 3. Your agent writes to content/drafts/

# 4. Review: give feedback OR approve
content-kit review content/drafts/my-post.md
# → Enter feedback → moves to reviewed/, notifies agent
# → No feedback → asks "Approve?" → moves to approved/

# 5. Agent revises (if feedback given), you review again

# 6. Post when approved
content-kit post content/approved/my-post.md

content-kit init automatically sets up ~/.content-kit.json with your workspace path, so commands work from any directory.

Content Folders

content/
├── drafts/        # Agent writes here
├── reviewed/      # You reviewed, awaiting agent revision
├── revised/       # Agent revised, ready for another look
├── approved/      # You approved, ready to post
└── posted/        # Archive after posting

The Workflow

┌─────────┐     ┌──────────┐     ┌─────────┐     ┌──────────┐     ┌────────┐
│ drafts/ │ ──▶ │ reviewed/│ ──▶ │ revised/│ ──▶ │ approved/│ ──▶ │ posted/│
└─────────┘     └──────────┘     └─────────┘     └──────────┘     └────────┘
   agent          human            agent           human           human
   writes         reviews          revises         approves        posts
                     ▲                │
                     └────────────────┘
                      more feedback

Secure Mode (Cryptographic Approval)

For extra assurance that content was human-approved, use --secure:

content-kit init --secure

This creates an Ed25519 signing keypair:

• Private key — encrypted with your password, stored in .content-kit-key
• Public key — embedded in the key file for verification

How it works:

1. When you approve content, you enter your password
2. The content is signed with your private key
3. When posting, the signature is verified
4. If content was modified after approval, posting is blocked

Why use it? Because you don't want to give the credentials to your social media to your AI agent. You can still automate posting in a boring deterministic process of course.

Files:

• .content-kit-key — your encrypted keypair (add to .gitignore!)
• Approved posts get approval_signature and content_hash in frontmatter

CLI Reference

# Setup
content-kit init              # Initialize content structure + global config
content-kit init --secure     # Also enable cryptographic approval signatures
content-kit auth <platform>   # Authenticate (linkedin, x)

# Workflow
content-kit list              # Show all folders with timestamps
content-kit review <file>     # Review: give feedback OR approve (if no feedback)
content-kit edit <file>       # Open in $EDITOR
content-kit post <file>       # Post (shows preview, asks confirmation)
content-kit post <file> -n    # Dry-run (--dry-run)

Platforms

LinkedIn

• Playwright browser automation
• Session encrypted in ~/.content-kit/

X (Twitter)

• Uses bird CLI
• Tokens extracted from Firefox, encrypted with password

OpenClaw Integration

If you're using OpenClaw, content-kit automatically notifies your agent when you give review feedback.

How it works:

1. content-kit init auto-detects OpenClaw and saves its path to .content-kit.json
2. When you run content-kit review <file> and enter feedback
3. The feedback is saved to the draft file
4. Your agent receives a message with the feedback and instructions to revise

The agent sees:

📝 Review feedback for 2025-01-30-linkedin-post.md:

"Make the intro punchier, less formal"

Read the draft at content/reviewed/..., apply the feedback, 
and save the revised version. Then confirm what you changed.

This creates a seamless review loop — you give feedback in terminal, agent responds in chat.

Manual config (if auto-detect fails):

{
  "clawdbotPath": "/path/to/clawdbot",
  "clawdbotTarget": "telegram:123456789"
}

For AI Agents

• ✅ Write to content/drafts/
• ✅ Move reviewed files to content/revised/
• ❌ Cannot approve or post

License

MIT — Lars de Ridder