code_architect_agent

Code Architect Agent (Go)

Go-based AI partner for:

• Code editing guidance (diff-first responses)
• New architecture/design discussions
• Migration and refactor planning

1) Configure environment

Create a local env file:

cp .env.example .env

This agent is configured for GitHub Models only.

GitHub Models

• GITHUB_MODELS_TOKEN (GitHub PAT)
• optional GITHUB_MODELS_MODEL (default: openai/gpt-5-mini)
• optional GITHUB_MODELS_ENDPOINT (default: https://models.github.ai/inference/chat/completions)

2) Run

From repo root:

go run ./agents/code_architect_agent

Server starts on PORT (default 8087).

Note: go run ./agents/code_architect_agent now auto-loads env values from either:

• .env (repo root)
• agents/code_architect_agent/.env

3) API

Health

curl http://localhost:8087/health

Chat

curl -X POST http://localhost:8087/chat \
  -H "Content-Type: application/json" \
  -d '{
    "mode": "architecture",
    "context": "Service currently monolithic with MongoDB",
    "message": "Propose target architecture for scaling and safer deployments"
  }'

mode can be:

• edit
• architecture
• general (default)

4) VS Code task

Run task: Run Agent HTTP Server from agents/code_architect_agent/.vscode/tasks.json.