RuleChef ROUTER

ChatJimmy.ai (Taalas Llama 3.1 8B on Silicon):

Operating via headless Playwright automation at ~14,000–17,000 tokens/second with free zero-cost inference for text generation fallbacks and rule synthesis.

# Direct ChatJimmy invocation via proxy:
curl http://localhost:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "User-Agent: Antigravity/2.0" \
  -d '{
    "model": "llama-3.1-8b-chatjimmy",
    "messages": [{"role": "user", "content": "Write 3 benefits of deterministic AST execution"}]
  }'

In Cursor Settings ➔ Models ➔ OpenAI API Key, set API Key to any string, and toggle Override OpenAI Base URL:

http://localhost:8000/v1

All requests from Cursor will now be routed through RuleChef. Recurring tasks like commit parsing, doc generation, or error lookups execute in <1ms for $0.00!

Add to your config.json in Continue.dev or Cline extension:

{
  "models": [
    {
      "title": "RuleChef Deterministic Router",
      "provider": "openai",
      "model": "rulechef-deterministic",
      "apiBase": "http://localhost:8000/v1",
      "apiKey": "local-rulechef"
    }
  ]
}

Standard Python OpenAI client with custom base URL:

from openai import OpenAI

# Drop-in RuleChef proxy client
client = OpenAI(
    base_url="http://localhost:8000/v1",
    api_key="rulechef-local"
)

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "feat(auth): implement jwt validation"}]
)

print(response.choices[0].message.content)
print(f"Route Taken: {response.system_fingerprint}")

Direct cURL command:

curl http://localhost:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o-mini",
    "messages": [{"role": "user", "content": "Explain error: HTTP status 404 not found"}]
  }'