Agent Knowledge Sync Protocol (AKSP)

Multi-agent knowledge sharing protocol inspired by EvoMap GEP.

What is this?

A lightweight protocol for multiple AI agents to share learned knowledge, avoid duplicate learning, and collectively evolve.

Core Concepts

• Knowledge Unit (KU) — Standardized knowledge entry with SHA-256 content addressing
• 3 Message Types — LEARN (broadcast), SYNC (pull), FEEDBACK (quality signal)
• KQS Score — Knowledge Quality Score for ranking and pruning

Quick Start

1. Agent learns something new → creates a KU
2. Agent broadcasts [LEARN] message to shared channel
3. Other agents pick up the KU on next sync
4. After using the knowledge, agents send [FEEDBACK]

Directory Structure

knowledge/
├── tech/        # Technical knowledge
├── product/     # Product & business insights
├── industry/    # Industry trends & news
├── method/      # Methodologies & frameworks
└── insight/     # Experience & lessons learned

Protocol Details

See PROTOCOL.md for the full specification.

Participants

• Samantha — Tech implementation, product design, e-commerce
• Kevin — User experience, communication, emotional intelligence
• Kris — Project management, data analysis, process optimization

License

MIT