Implement Phase 2 - Test Server & Multi-Agent Coordination

[yusufkaraaslan]

Phase 2: Test Server & Multi-Agent Coordination

📋 Overview

Build a framework-agnostic Test Server that coordinates test execution across multiple agents, preparing the foundation for Phase 4's multi-agent support.

Status: Phase 1 complete ✅ | Multi-framework support complete ✅
Goal: Enable safe concurrent development by coordinating test execution

🎯 Objectives

1. Test Server - HTTP API that queues and executes tests sequentially
2. Framework-Agnostic Design - Works with all 15+ supported frameworks
3. Agent Coordination - Prevents test conflicts when multiple agents run
4. Enhanced Agent Runner - Integration with Test Server
5. Preparation for Multi-Agent - Foundation for Phase 4

🔑 Why Phase 2 is Critical

Problem: When multiple agents finish work simultaneously, they all try to run tests at once:

• Only one test process can run per project at a time
• Parallel test execution causes resource conflicts
• Results can interfere with each other

Solution: Test Server acts as a coordinator:

Agent 1 ──┐
Agent 2 ──┼──> Test Server (Queue) ──> Single Test Process
Agent 3 ──┘     (Async API)              (Sequential)

📦 Components to Implement

1. Test Server (scripts/test-server.py)

API Endpoints:

• POST /test/submit - Submit test job
• GET /test/status/<job_id> - Check status
• GET /test/results/<job_id> - Get results
• GET /health - Health check
• GET /queue - View queue

2. Enhanced Agent Runner

• Submit tests to server instead of running directly
• Poll for results
• Retry logic
• Only create PR if tests pass

3. Systemd Service

• systemd/test-server.service
• Auto-restart, logging

4. Multi-Framework Support

Works with all 15+ frameworks via presets

✅ Acceptance Criteria

• Test Server runs as systemd service
• Queues and executes tests sequentially
• Works with 4+ frameworks (Godot, React, Django, Rust)
• Agent runner integrates with server
• PRs only created if tests pass
• Complete documentation

📝 Implementation Plan

Week 1: Core Server (Flask API, queue, execution)
Week 2: Integration (agent-runner, multi-framework testing, docs)

Estimated: 2 weeks full-time, 3-4 weeks part-time

---

Complexity: Medium-High
Priority: High (Required for Phase 4)