Archon Architecture POC - Three-Layer Implementation #1

ezra · 2026-04-02T19:50:23Z

ezra commented

2026-04-02 19:50:23 +00:00

Overview

Implements Epic #370: Archon Architecture proof-of-concept.

Three-Layer Architecture

Layer	Component	Purpose
1	Thin Hermes Profile	Identity/routing only (< 50 lines)
2	Claw Code Runtime	All business logic, tool registry
3	Gemma via Ollama	Local intelligence at localhost:11434

Files Added

archon-poc/README.md - Documentation
archon-poc/profile.yaml - Thin Hermes layer (26 lines)
archon-poc/runtime/harness.py - Claw Code stub
archon-poc/runtime/tool_registry.py - Tool definitions
archon-poc/ollama_client.py - Gemma 4 interface
archon-poc/test_integration.py - End-to-end test (12 tests)
archon-poc/.gitignore - Python artifacts

Features

Tool registry with pattern matching (/time, /status, /echo, /ollama_list)
Full routing between tool execution and intelligence layer
Ollama client with streaming and chat support
Connection to localhost:11434 using gemma3:4b model

Testing

All 12 integration tests pass:

python3 test_integration.py

Key Requirements Met

profile.yaml is THIN - only identity, constraints, routing
No reasoning logic in profile - everything goes to Claw runtime
Connects to Ollama at localhost:11434
Uses gemma3:4b model (gemma4:4b when available)
Integration test verifies full flow works

Dispatch Reference

Closes Dispatch #371

## Overview Implements Epic #370: Archon Architecture proof-of-concept. ## Three-Layer Architecture | Layer | Component | Purpose | |-------|-----------|---------| | 1 | Thin Hermes Profile | Identity/routing only (< 50 lines) | | 2 | Claw Code Runtime | All business logic, tool registry | | 3 | Gemma via Ollama | Local intelligence at localhost:11434 | ## Files Added - `archon-poc/README.md` - Documentation - `archon-poc/profile.yaml` - Thin Hermes layer (26 lines) - `archon-poc/runtime/harness.py` - Claw Code stub - `archon-poc/runtime/tool_registry.py` - Tool definitions - `archon-poc/ollama_client.py` - Gemma 4 interface - `archon-poc/test_integration.py` - End-to-end test (12 tests) - `archon-poc/.gitignore` - Python artifacts ## Features - Tool registry with pattern matching (`/time`, `/status`, `/echo`, `/ollama_list`) - Full routing between tool execution and intelligence layer - Ollama client with streaming and chat support - Connection to localhost:11434 using gemma3:4b model ## Testing All 12 integration tests pass: ``` python3 test_integration.py ``` ## Key Requirements Met - profile.yaml is THIN - only identity, constraints, routing - No reasoning logic in profile - everything goes to Claw runtime - Connects to Ollama at localhost:11434 - Uses gemma3:4b model (gemma4:4b when available) - Integration test verifies full flow works ## Dispatch Reference Closes Dispatch #371

ezra added 2 commits 2026-04-02 19:50:23 +00:00

Implement Archon Architecture POC (Epic #370 , Dispatch #371 ) 1b3bca9902

Three-layer architecture implementation:
- Layer 1: Thin Hermes profile (profile.yaml, < 50 lines)
- Layer 2: Claw Code runtime (harness.py, tool_registry.py)
- Layer 3: Gemma via Ollama (ollama_client.py)

Features:
- Tool registry with pattern matching (/time, /status, /echo, /ollama_list)
- Full routing between tool execution and intelligence layer
- Ollama client with streaming and chat support
- Comprehensive integration test suite (12 tests)
- Connection to localhost:11434 using gemma3:4b model

Add .gitignore for Python artifacts 19fae5a6e5

ezra referenced this pull request from Timmy_Foundation/timmy-home

2026-04-02 19:52:02 +00:00

[BURN-DOWN] Complete Architecture Proof-of-Concept #374

ezra merged commit 7711b7e279 into master

2026-04-04 12:17:03 +00:00

ezra referenced this issue from a commit

2026-04-04 12:17:05 +00:00

Merge pull request 'Archon Architecture POC - Three-Layer Implementation' (#1) from feature/archon-poc-implementation into master

Sign in to join this conversation.