Refactor Predictor and market data handling

- Added Predictor class with input preparation and instructions for financial strategy generation.
- Removed PredictorAgent class and integrated its functionality into the new Predictor module.
- Created a base market API wrapper and specific implementations for Coinbase and CryptoCompare.
- Introduced PublicBinanceAgent for fetching public prices from Binance.
- Refactored ToolAgent to utilize the new Predictor and market API wrappers for improved data handling and predictions.
- Updated models to streamline the selection of available LLM providers.
- Removed deprecated signer classes for Coinbase and CryptoCompare.

src/app/tool.py

@@ -1,27 +1,36 @@
-from app.agents.market_agent import MarketAgent
 from app.agents.news_agent import NewsAgent
 from app.agents.social_agent import SocialAgent
-from app.agents.predictor_agent import PredictorAgent
+from app.agents import predictor
+from app.agents.predictor import PredictorStyle
+from app.markets import get_first_available_market_api
 from app.models import Models
 
 
 class ToolAgent:
-    def __init__(self, available_models: list[Models]):
-        self.market_agent = MarketAgent()
-        self.news_agent = NewsAgent()
-        self.social_agent = SocialAgent()
-        self.predictor_agent = PredictorAgent()
+    def __init__(self, available_models: list[Models], all_styles: list[PredictorStyle]):
+        self.available_models = available_models
+        self.all_styles = all_styles
 
-    def interact(self, query: str, provider: str, style: str):
-        """
-        Funzione principale che coordina gli agenti per rispondere alla richiesta dell'utente.
-        """
+        self.market = get_first_available_market_api(currency="USD")
+        self.choose_provider(0) # Default to the first model
+
+    def choose_provider(self, index: int):
         # TODO Utilizzare AGNO per gestire i modelli... è molto più semplice e permette di cambiare modello facilmente
         # TODO https://docs.agno.com/introduction
         # Inoltre permette di creare dei team e workflow di agenti più facilmente
+        chosen_model = self.available_models[index]
+        self.predictor = chosen_model.get_agent(predictor.instructions())
+        self.news_agent = NewsAgent()
+        self.social_agent = SocialAgent()
+
+    def interact(self, query: str, style_index: int):
+        """
+        Funzione principale che coordina gli agenti per rispondere alla richiesta dell'utente.
+        """
 
         # Step 1: raccolta analisi
-        market_data = self.market_agent.analyze(query)
+        cryptos = ["BTC", "ETH", "XRP", "LTC", "BCH"]  # TODO rendere dinamico in futuro
+        market_data = self.market.get_products(cryptos)
         news_sentiment = self.news_agent.analyze(query)
         social_sentiment = self.social_agent.analyze(query)
 
@@ -29,11 +38,15 @@ class ToolAgent:
         sentiment = f"{news_sentiment}\n{social_sentiment}"
 
         # Step 3: previsione
-        prediction = self.predictor_agent.predict(
+        inputs = predictor.prepare_inputs(
             data=market_data,
-            sentiment=sentiment,
-            style=style,
-            provider=provider
+            style=self.all_styles[style_index],
+            sentiment=sentiment
         )
 
-        return f"{market_data}\n{sentiment}\n\n📈 Consiglio finale:\n{prediction}"
+        prediction = self.predictor.run(inputs)
+        #output = prediction.content.split("</think>")[-1] # remove thinking steps and reasoning from the final output
+        output = prediction.content
+
+        market_data = "\n".join([f"{product.symbol}: {product.price}" for product in market_data])
+        return f"{market_data}\n{sentiment}\n\n📈 Consiglio finale:\n{output}"