A human-in-the-loop content automation workflow for AI agents to draft, revise, and stage social media posts securely.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agent-content-pipeline
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install agent-content-pipeline using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Agent Content Pipeline?
The Agent Content Pipeline is a specialized framework designed to bridge the gap between AI-generated drafts and final publication. By implementing a strict state-based directory structure, it allows AI agents to contribute to the creative process while maintaining absolute human control over the final output. This skill ensures that no content is published without explicit human intervention, making it an essential tool for maintaining brand voice and operational security in automated environments.
By leveraging Openclaw Skills, this pipeline organizes content into logical lifecycle stages: drafts, reviewed, revised, approved, and posted. It includes built-in platform guidelines for LinkedIn, X (Twitter), and Reddit, ensuring that AI-generated text adheres to idiomatic standards, character limits, and technical requirements like YAML frontmatter metadata.
Agent Content Pipeline Use Cases
- Automating the drafting process for multi-platform social media campaigns.
- Implementing a rigorous peer-review system between an AI agent and a human editor.
- Managing content versioning and feedback threads for technical updates or marketing posts.
- Securely staging posts for LinkedIn and X with encrypted authentication and human approval triggers.
How Agent Content Pipeline Works
- The AI agent generates a post file with specific YAML frontmatter and saves it in the drafts directory.
- A human reviewer utilizes the CLI to examine the draft and provide specific feedback, which moves the file to the reviewed folder.
- The agent reads the human feedback, applies the necessary revisions, and moves the file to the revised directory for a second look.
- Once the human is satisfied, they manually approve the content via the CLI, which moves it to the approved stage.
- The human executes the final post command to publish the content to the designated social platform, ensuring no automated accounts post without permission.
Agent Content Pipeline Setup
To integrate this workflow into your Openclaw Skills environment, install the package and initialize your project directory:
npm install -g agent-content-pipeline
content init . # Creates folders and global configuration
For high-security environments requiring cryptographic approval signatures, use the secure initialization flag:
content init . --secure
Agent Content Pipeline Data Schema & Taxonomy
The skill utilizes a structured file-based system for state management, using the naming convention YYYY-MM-DD-<platform>-<slug>.md. Each file includes YAML frontmatter to track status and platform targets.
| Directory | Description |
|---|---|
drafts/ |
Initial work in progress created by the agent. |
reviewed/ |
Human-reviewed files awaiting agent revision. |
revised/ |
Agent-updated content ready for final approval. |
approved/ |
Verified content staged for publication. |
posted/ |
Archive of successfully published content. |
.content-pipeline/threads/ |
Internal logs for feedback and revision history. |
Agent Content Pipeline Advanced Features
- Cryptographic approval signatures to prevent unauthorized content movement.
- Platform-specific validation for LinkedIn CTAs and X character constraints.
- Manual cookie extraction and encrypted token storage for secure social media authentication.
- Dry-run functionality to preview post rendering before live deployment.
- Support for feedback threads to maintain a history of agent-human collaboration.
SKILL.md
Loading
Related Openclaw Skills
Cookidoo Thermomix Meal Planner
A robust Python-based CLI skill for managing Cookidoo meal plans, recipes, and shopping lists through an AI agent.
Munich Public Transport (MVG) CLI
A terminal-based interface for accessing real-time Munich public transport data and live S-Bahn tracking.
Knuspr CLI
A powerful CLI-based automation skill for managing grocery shopping, cart items, and delivery slots on Knuspr.de.
apo-cli
A Python-based CLI skill for searching pharmacy products and managing shopping carts on apohealth.de without authentication.
Obsidian Task Management
A comprehensive task management framework for Obsidian vaults featuring Kanban boards, Dataview dashboards, and YAML-driven task notes.
Obsidian Tasks
A comprehensive task management framework for Obsidian vaults that integrates Kanban boards, Dataview dashboards, and structured YAML metadata.
Featured*
