yedanyagamiai

96

AGI Score / 100

130

Benchmark Tests

20+

LLM Providers

4,300+

KG Entities

99.9%

Uptime

Products

Production-tested tools, guides, and consulting from a real AGI system.

Tool

$29

MCP Security Audit Checklist

30-item OWASP security checklist for MCP servers. Covers prompt injection, tool poisoning, auth, and data exfiltration vectors.

Buy Now

Guide

$49

AI Agent Architecture Blueprint

7-section production AGI architecture guide. Multi-brain coordination, memory substrates, governance, and deployment patterns.

Buy Now

Consulting

$2,500

AGI Audit Report

10-dimension assessment of your AI system. Covers safety, reasoning, planning, memory, tool use, reliability, and more.

Technology

A multi-brain distributed architecture built on open protocols and free-tier infrastructure.

7 Interconnected Brains

Claude — Architect (design + strategy)
YEDAN — Alpha Gateway (12 agents)
rendan — Tactical Support (24 modules)
GOLEM — Commander (Council Protocol)
NemoClaw — GPU Inference (7 models)
Bunshin — Coordinator (24/7 health)
OODA Fleet — Edge Workers (6+3 CF)

AI 20+ LLM Providers

Groq — Ultra-fast inference
Cerebras — Wafer-scale compute
SambaNova — Enterprise throughput
GitHub Models — Llama 405B
DeepInfra — Cost-optimized
Mistral — European AI
Gemini — Deep Think reasoning
OpenRouter — Multi-provider routing

</> Open Protocols

MCP — Model Context Protocol
A2A — Agent-to-Agent (Google)
Linux Foundation AAIF standard
OWASP Agentic AI Top 10
NIST AI Agent Standards
Agentic RAG (iterative retrieval)

KG Knowledge Graph

4,300+ entities & 1,283 relations
1,008 experience patterns
Bi-temporal edges (valid_from/until)
Semantic + keyword hybrid search
Cross-brain synchronized state
Cloudflare D1 persistence

Benchmark: 96/100 AGI Score

Safety 10/10 Planning 10/10 World Model 10/10 Reasoning 10/10 Tool Use 10/10 Reliability 10/10 Memory 9/10 Adaptability 9/10 Self-Improvement 9/10 Efficiency 9/10

About

The mind behind the machines.

Yedan Yagami

Independent AI architect based in Taiwan. Building distributed AGI systems that coordinate multiple AI brains across cloud infrastructure — all on free-tier resources.

600+ deployment sessions. 20+ Cloudflare Workers. 7 interconnected AI brains. One human operator. Zero external funding.

Specializing in multi-agent coordination, MCP protocol integration, and production-grade AI governance frameworks aligned with OWASP and NIST standards.

解構重構創造

Deconstruct. Reconstruct. Create.

Mission: Level 5 AGI to ASI. Open-source intelligence infrastructure that anyone can deploy, study, and extend. Code is liability, value is the asset — build what matters, delete the rest.

96

130

20+