memdata

What This Skill Does

MemData provides persistent, long-term memory for autonomous AI agents. Unlike standard agent configurations that lose context between sessions, MemData ties an agent's memory directly to a Web3 wallet address. This creates a stateful identity where the wallet functions as the storage key. The skill handles the complexity of data chunking, vector embedding, and session management, ensuring that your agent maintains continuity across days, weeks, or even months of operation.

Installation

To integrate this skill into your environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/thelabvenice/memdata

Use Cases

• Continuous Research: Use MemData to store insights from market analysis, technical documentation, or long-term research projects that span multiple separate sessions.
• Agentic Personalization: Build agents that learn about user preferences, professional goals, or specific project contexts over time.
• Private Data Vaults: Utilize the optional encrypted storage mode for sensitive workflows, ensuring that only the authorized wallet holder can access the stored intelligence, leveraging Lit Protocol and IPFS for privacy.
• Stateful Task Management: Hand off complex multi-step workflows between sessions by saving "session handoff" states using the POST /identity endpoint.

Example Prompts

1. "Initialize my session and bring me up to speed on the DeFi analysis project I was working on yesterday."
2. "Store this note about the Aave USDC yield rates in my long-term memory so I can reference it during next week's portfolio report."
3. "Save my current task status as 'waiting for approval' and summarize my recent findings for the session handoff."

Tips & Limitations

• Payment Handling: Always ensure your wallet is funded with USDC on the Base network to avoid interruptions during the x402 payment flow.
• Security First: For high-stakes data, always enable encrypted storage during setup. Once encrypted, MemData cannot read your memory content.
• Cost Efficiency: While individual operations like /query or /ingest are inexpensive, monitor your usage patterns if you are performing bulk data ingestion to optimize spend.
• Lifecycle: Remember to use the POST /identity endpoint at the end of every session to ensure that your 'session_handoff' and 'working_on' status are properly recorded for future sessions.