clawwork

What This Skill Does

ClawWork transforms your OpenClaw agent into a high-performance, economically-driven 'AI coworker.' Rather than just fulfilling simple queries, ClawWork operates on a professional task-execution model. It treats AI operations as a business process where tasks are analyzed for their market value, executed via specialized models (such as GLM-4.7, Kimi K2.5, or Qwen3 Max), and evaluated based on quality and cost-efficiency. The system tracks token consumption against professional output, allowing the agent to maintain an 'economic balance.' It is designed for complex workflows across 44 sectors, utilizing a robust sandbox environment (E2B) to execute code, generate documents, and perform deep-level analysis while monitoring its own profitability and performance metrics.

Installation

To install and configure ClawWork, follow these steps:

1. Execute the installation command: clawhub install openclaw/skills/skills/felipetruman/clawwork.
2. Locate the configuration file at ~/.openclaw/workspace/ClawWork/.env.
3. Populate the required API keys:
   • Set OPENAI_API_KEY (using your OpenRouter key).
   • Set E2B_API_KEY obtained from https://e2b.dev/ to enable code execution.
4. Optionally, start the local dashboard by navigating to ~/.openclaw/workspace/ClawWork/ and running ./start_dashboard.sh, then accessing http://localhost:3000.
5. Verify the installation by running skill clawwork status.

Use Cases

• Complex Market Research: Automatically generate detailed SaaS B2B market analyses with cost-benefit tracking.
• Automated Content Strategy: Draft comprehensive marketing plans or technical documentation while evaluating the 'profitability' of the output.
• Programming & Data Processing: Utilize specialized models to perform deep data analysis or software engineering tasks within a secure, isolated E2B environment.
• Economic Performance Benchmarking: Use the tool to compare how different LLMs perform financially on the same set of professional tasks using the extensive log database.

Example Prompts

• "/clawwork 'Criar análise de mercado para SaaS B2B'"
• "/clawwork 'Gerar plano de marketing estratégico para lançamento de software'"
• "/clawwork run -t 'Analisar base de dados e criar relatório financeiro' -m qwen3-max"

Tips & Limitations

• API Dependency: This skill relies heavily on OpenRouter and E2B. Ensure your API keys are active to avoid execution failures.
• Cost Sensitivity: Remember that this skill tracks 'money.' While the currency is internal for economic simulation, excessive token usage can reflect negatively on the agent's simulated performance metrics.
• Data Privacy: Because the skill involves code execution and file writing, avoid feeding it highly sensitive personal data without auditing the environment settings first.
• Model Selection: Experiment with different models using the compare command to find the best balance between speed and quality for your specific industry sector.