vault

What This Skill Does

The Vault skill provides a robust, local orchestration engine designed to manage the lifecycle of long-running research tasks within the OpenClaw AI environment. By utilizing a persistent local SQLite database located at ~/.researchvault/, it ensures that all research states, logs, and findings are preserved across sessions without reliance on expensive external APIs or cloud-based state management. It excels at bridging the gap between raw data collection and structured insight generation, allowing users to track project objectives, log individual research events, and maintain high levels of analytical confidence through its integrated metadata tracking system. By keeping all research data localized, it guarantees privacy and offline availability, making it an essential tool for deep-dive investigations.

Installation

To integrate this skill into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/lraivisto/researchvault-brain Once installed, ensure your environment has the necessary write permissions to the ~/.researchvault/ directory. If you prefer to change the default storage path, you may define the RESEARCHVAULT_DB environment variable prior to usage.

Use Cases

• Project Tracking: Ideal for multi-stage research projects where progress needs to be saved and reviewed over days or weeks.
• Evidence Management: Effectively curate sources and document individual research steps, including tagging them with confidence levels for better auditability.
• Structured Output: Automatically generate comprehensive project reports and summaries in markdown format for easy sharing or documentation.
• Data Integrity: Securely manage research states locally, ensuring that no sensitive data is leaked to external third-party servers during the research process.

Example Prompts

1. "Initialize a new project named 'quantum-leaps' with the objective of summarizing the latest findings in quantum computing hardware."
2. "Use the scuttle command to scrape the contents of https://arxiv.org/list/quant-ph/recent for my 'quantum-leaps' project."
3. "Export all findings and research summaries for the 'quantum-leaps' project into a clean markdown file named 'quantum_report.md'."

Tips & Limitations

• Database Maintenance: Since the database is stored locally, it is automatically excluded from your version control systems (like Git). If you move machines, you will need to back up the SQLite database file manually.
• Confidence Scores: Always assign confidence scores when logging new research; this helps you filter for high-quality information when generating your final project summaries.
• SSRF Security: The scuttle function includes built-in SSRF (Server-Side Request Forgery) protection, making it safe to use on various URLs, but always exercise caution when directing automated tools toward unknown or untrusted web resources.