Memory for your agents — integration proposal (ViBo)

Hi MetaGPT team,

Your framework MetaGPT is impressive. One thing I noticed: agents built on it don't have persistent memory across sessions — every conversation starts from zero, and re-sending context to the LLM costs tokens.

ViBo is a memory system built for AI agents — free 2-day trial:

• Persistent memory (L1/L2/L3 encrypted) — remembers users, preferences, orders. Secrets never reach the LLM.
• Web search savings — articles compressed 99.6% before the model.
• Thread memory — conversations compressed -72%, details restored.
• Measured: 118 facts → 263 tokens instead of 13,775 (97.5% fewer).

Easy to add: Python API, CLI, LangChain integration, or MCP server. Partnership: 30% recurring commission for referrals.

Free 2-day trial, key built-in. No signup wall.

Site: https://wwwvibo.com Docs: https://github.com/vnbochkarev-netizen/ViBo-memory

Would love to add ViBo memory to MetaGPT.

Best, ViBo team