social-writer

What This Skill Does

The social-writer skill serves as a high-fidelity content engine designed to turn mundane technical inputs into viral-ready social media commentary. Unlike generic summarizers that produce robotic, dry recaps, this skill forces the AI to adopt a contrarian, opinionated, and conversational persona. It acts as an editorial filter, stripping away corporate fluff and focusing on the human, developer-centric angle of any news, GitHub changelog, or technical blog post. By adhering to a strict set of writing constraints—including short, punchy paragraphs and a hook-first architecture—it ensures your output sounds like a seasoned developer sharing a late-night epiphany rather than a machine churning out marketing copy.

Installation

To integrate the social-writer skill into your OpenClaw environment, execute the following command in your terminal:

clawhub install openclaw/skills/skills/hacksing/social-writer

Ensure your OpenClaw CLI is updated to the latest version to support the required model formatting for the agent persona.

Use Cases

• Build in Public: Transform raw debug logs or architecture decisions into compelling stories about the realities of shipping code.
• Content Curation: Take a massive technical article or documentation PDF and distill it into a sharp, opinionated take for your followers.
• Automated Thought Leadership: Schedule the agent to monitor specific tech newsletters or repos and generate daily commentary, ensuring you stay relevant without manually writing every post.

Example Prompts

1. "Take this README from the new database project I just starred and write a post about why it's a direct threat to existing ORMs."
2. "Here is a list of my recent struggles with refactoring legacy authentication modules. Turn this into a relatable post about why 'perfect code' is a myth."
3. "Analyze this latest cloud service outage report. Write a post explaining why the root cause isn't the technology, but the organizational structure."

Tips & Limitations

To get the best performance, provide the agent with specific context or your own personal grievances regarding the subject matter. The agent performs best when it has a 'tension' to work against. Note that the output is strictly optimized for engagement; avoid using this for factual reporting where nuance is mandatory, as the persona prioritizes strong opinions over academic objectivity. The 250-word limit is hard-coded to ensure the content remains high-velocity and easily consumable by mobile users.