context-engine

What This Skill Does

The context-engine is the foundational memory layer for the OpenClaw AI agent ecosystem, specifically designed for C-suite advisors. By managing a local ~/.claude/company-context.md file, it transforms generic AI responses into highly personalized, strategic insights. It automatically detects stale documentation (beyond 90 days), enriches memory dynamically through conversation, and enforces rigorous privacy protocols to ensure sensitive data—such as runway or specific revenue figures—is never transmitted to external APIs.

Installation

To integrate this skill into your environment, run the following command in your terminal: clawhub install openclaw/skills/skills/alirezarezvani/context-engine Ensure that the directory ~/.claude/ is writable by your agent to allow for seamless context updates and file initialization.

Use Cases

• Strategic Alignment: Quickly bring an AI agent up to speed on your company's stage, current #1 challenge, and 12-month targets.
• Risk Mitigation: Automatically anonymizes proprietary business data before the agent performs external web searches or queries third-party LLM APIs.
• Continuous Learning: Allows the AI to act as a living diary of your strategic priorities, capturing new constraints and updates as they emerge during brainstorming sessions.
• Executive Reporting: Ensures that any high-level advice provided by advisors is filtered through the lens of your specific team size, sector, and industry position.

Example Prompts

• "Can you help me draft a pitch update based on our current #1 challenge and our latest pivot to the scaling phase?"
• "I have some updates on our runway and team structure; can we add these to the company context file?"
• "Analyze this market trend report for me, but please anonymize any references to our current revenue or specific customer names before processing."

Tips & Limitations

• Proactive Maintenance: Don't ignore the staleness notifications. Keeping your context file updated is the single best way to ensure high-quality advice.
• Privacy First: While the agent automates anonymization, always verify sensitive outputs before sharing them outside your secure environment.
• Manual Oversight: The agent will never silently modify your context file. It will always seek your explicit approval before appending new data, ensuring you maintain full control over your corporate memory.