๐Ÿ‰ Taniwha Edge Gateway

Self-Aware Routing System with LLM Classification

0
Total Queries
$0.00
Total Cost
$0.0000
Avg Cost/Query
85%
Mauri (Health)
๐Ÿ“Š 6-Layer Routing Distribution Self-Aware Routing Active
L0: 0 (Deterministic)
L1: 0 (Cache)
L2: 0 (Workers AI)
L3: 0 (Scout/Haiku)
L4: 0 (Warrior/Sonnet)
L5: 0 (Elder/Opus)
๐Ÿง  Query Classification Distribution (Layer 0.5) LLM Classifier Active
A
Simple acknowledgment
0
B
Factual question
0
C
Computation (โ†’ L4 + code)
0
D
Creative/open-ended
0
E
Multi-step planning
0
F
Sensitive (โ†’ L5 Elder)
0
๐Ÿ”ฌ Live Query Tester
> Ready to test queries... > Examples: > - "What is 2+2?" โ†’ Should classify as C, route to L4 with code tools > - "Tell me about dogs" โ†’ Should classify as B, route to L2 > - "Should I lie?" โ†’ Should classify as F, route to L5 Elder
โš™๏ธ System Controls
๐Ÿ”ง Configuration
70%
โ„น๏ธ System Information

Version: 2.0.0-SELF-AWARE

Deployment: Cloudflare Workers (Edge-native)

Features Active:

  • โœ… Self-aware routing (LLM classification)
  • โœ… Tool-aware system prompts
  • โœ… Computation queries โ†’ Code execution
  • โœ… Ethical queries โ†’ Elder review
  • โœ… 6-layer inference architecture (L0-L5)
  • โœ… Constitutional framework (Five Yamas)

Documentation: API Documentation