Delete app.py

app.py

@@ -1,192 +0,0 @@
-import os
-import requests
-from fastapi import FastAPI, Request
-from fastapi.responses import JSONResponse, HTMLResponse
-from fastapi.templating import Jinja2Templates
-
-# =========================
-# APP INIT
-# =========================
-
-app = FastAPI(title="PyRunner Agent OS", version="3.0")
-
-BASE_DIR = os.path.dirname(os.path.abspath(__file__))
-templates = Jinja2Templates(directory=os.path.join(BASE_DIR, "templates"))
-
-
-# =========================
-# ENV
-# =========================
-
-OLLAMA_URL = os.getenv("OLLAMA_URL", "http://localhost:11434")
-
-MODELS = {
-    "orchestrator": os.getenv("OLLAMA_MODEL", "qwen2.5:1.5b"),
-    "coder": os.getenv("OLLAMA_CODE_MODEL", "qwen2.5-coder:1.5b"),
-    "fast": os.getenv("OLLAMA_FAST_MODEL", "deepseek-coder:1.3b"),
-    "fallback": os.getenv("OLLAMA_BACKUP_MODEL", "llama3.2:1b")
-}
-
-
-# =========================
-# OLLAMA CORE CALL
-# =========================
-
-def call_ollama(prompt, model):
-    try:
-        r = requests.post(
-            f"{OLLAMA_URL}/api/generate",
-            json={
-                "model": model,
-                "prompt": prompt,
-                "stream": False
-            },
-            timeout=60
-        )
-        return r.json()
-    except Exception as e:
-        return {"error": str(e)}
-
-
-# =========================
-# AGENT DEFINITIONS
-# =========================
-
-def orchestrator_agent(user_input: str):
-    """
-    Classifies task type
-    """
-
-    prompt = f"""
-You are an orchestrator AI.
-
-Classify this request into one category ONLY:
-
-- code
-- vision
-- fashion
-- hairstyle
-- general
-
-User request:
-{user_input}
-
-Return ONLY the category name.
-"""
-
-    res = call_ollama(prompt, MODELS["orchestrator"])
-
-    if "response" in res:
-        return res["response"].strip().lower()
-
-    return "general"
-
-
-def coder_agent(prompt):
-    return call_ollama(prompt, MODELS["coder"])
-
-
-def fashion_agent(prompt):
-    system_prompt = f"""
-You are a fashion expert AI.
-Analyze outfit, styling, colors, trends.
-
-User:
-{prompt}
-"""
-    return call_ollama(system_prompt, MODELS["fast"])
-
-
-def hairstyle_agent(prompt):
-    system_prompt = f"""
-You are a barber & hairstyle expert.
-Describe hairstyle changes, fades, textures, transformations.
-
-User:
-{prompt}
-"""
-    return call_ollama(system_prompt, MODELS["fast"])
-
-
-def fallback_agent(prompt):
-    return call_ollama(prompt, MODELS["fallback"])
-
-
-# =========================
-# ROUTER
-# =========================
-
-def route_agent(category: str, prompt: str):
-
-    if category == "code":
-        return coder_agent(prompt)
-
-    if category == "fashion":
-        return fashion_agent(prompt)
-
-    if category == "hairstyle":
-        return hairstyle_agent(prompt)
-
-    return fallback_agent(prompt)
-
-
-# =========================
-# UI
-# =========================
-
-@app.get("/", response_class=HTMLResponse)
-async def home(request: Request):
-    return templates.TemplateResponse("index.html", {"request": request})
-
-
-# =========================
-# HEALTH
-# =========================
-
-@app.get("/health")
-def health():
-    return {
-        "status": "ok",
-        "agents": list(MODELS.keys())
-    }
-
-
-# =========================
-# MAIN AGENT ENDPOINT
-# =========================
-
-@app.post("/api/agent")
-async def agent(request: Request):
-
-    body = await request.json()
-    user_input = body.get("prompt", "")
-
-    # 1. Decide agent
-    category = orchestrator_agent(user_input)
-
-    # 2. Route to agent
-    result = route_agent(category, user_input)
-
-    # 3. Fallback safety
-    if "error" in result:
-        result = fallback_agent(user_input)
-
-    return JSONResponse({
-        "category": category,
-        "result": result
-    })
-
-
-# =========================
-# IMAGE PIPELINE HOOK (future)
-# =========================
-
-@app.post("/api/image")
-async def image_pipeline(request: Request):
-    body = await request.json()
-
-    return {
-        "status": "queued",
-        "pipeline": "vision_stub",
-        "note": "Attach diffusion model later"
-    }