social-media-agent

What This Skill Does

The social-media-agent is a comprehensive AI-powered assistant designed to manage your entire social media presence. Rather than simply generating text, this skill acts as a holistic manager that plans, writes, schedules, and analyzes content across major platforms including X (Twitter), LinkedIn, Instagram, TikTok, Facebook, and Pinterest. By integrating with robust scheduling backends like Buffer or Postiz, it eliminates the manual friction of posting content, allowing you to focus on strategy and growth.

Installation

To integrate this skill into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/batsirai/social-media-engine

Ensure that you have completed the initial setup by verifying your scheduler backend with node scripts/post-scheduler.js --status. You will need either a Buffer API key or a self-hosted Postiz instance configured.

Use Cases

• Content Planning: Generate a consistent, strategic content calendar using the 70/20/10 rule (Value/Personality/Promotion) to ensure your feed remains engaging and avoids being overly promotional.
• Multi-Platform Repurposing: Take a single core idea or article and adapt it into platform-specific formats—short, punchy threads for X, professional insights for LinkedIn, or visual captions for Instagram.
• Automated Scheduling: Offload the manual labor of posting at specific times by drafting content within the agent and pushing it directly to your scheduling queue.
• Brand Voice Maintenance: Maintain a human-like, conversational tone across all channels, avoiding the generic corporate jargon that often plagues automated social media tools.

Example Prompts

1. "I need a content plan for next week. Our brand is a productivity SaaS, our audience is solopreneurs, and we have a new feature launch on Thursday."
2. "Write a 3-part thread for X about why we removed our welcome email sequence. Keep it opinionated, conversational, and avoid corporate buzzwords."
3. "Review my drafted posts for this week and tell me if they follow the 70/20/10 rule. If not, suggest some personality-driven tweaks."

Tips & Limitations

• Quality Control: Always review the generated output. While the agent follows specific stylistic guidelines to sound like a human, high-stakes announcements should still be proofread by a human.
• Platform Nuance: The agent is trained to avoid 'hashtag walls' and excessive self-promotion. If you find the agent is too opinionated or not formal enough, adjust your initial prompt to set a specific 'brand persona' constraint.
• Backend Dependency: The agent is only as reliable as your chosen backend (Buffer/Postiz). Ensure your API keys are kept secure and that your schedules are monitored periodically in your chosen dashboard.