Use PMAT from Pypi

main.py

@@ -3,9 +3,7 @@
 import typer
 from prediction_market_agent_tooling.markets.markets import (
     MarketType,
-    get_bet_amount,
     get_binary_markets,
-    place_bet,
 )
 
 import prediction_market_agent as pma
@@ -21,7 +19,7 @@ def main(
     Picks one market and answers it, optionally placing a bet.
     """
     # Pick a market
-    market = get_binary_markets(market_type)[0].to_agent_market()
+    market = get_binary_markets(market_type)[0]
 
     # Create the agent and run it
     agent = get_agent(agent_type)
@@ -44,13 +42,11 @@ def main(
             f"Placing bet with position {pma.utils.parse_result_to_str(result)} on market '{market.question}'"
         )
         amount = Decimal(
-            input(f"How much do you want to bet? (in {market.bet_amount_currency}): ")
+            input(f"How much do you want to bet? (in {market.currency}): ")
         )
-        place_bet(
-            market=market.original_market,
-            amount=get_bet_amount(amount, market_type),
+        market.place_bet(
+            amount=market.get_bet_amount(amount),
             outcome=result,
-            omen_auto_deposit=True,
         )
 
 

prediction_market_agent/agents/abstract.py

@@ -1,4 +1,4 @@
-from prediction_market_agent_tooling.markets.data_models import AgentMarket
+from prediction_market_agent_tooling.markets.agent_market import AgentMarket
 
 
 class AbstractAgent:

prediction_market_agent/agents/always_yes.py

@@ -1,4 +1,4 @@
-from prediction_market_agent_tooling.markets.data_models import AgentMarket
+from prediction_market_agent_tooling.markets.agent_market import AgentMarket
 
 from prediction_market_agent.agents.abstract import AbstractAgent
 

prediction_market_agent/agents/autogen_agent.py

@@ -3,7 +3,7 @@
 
 import autogen
 from autogen.agentchat.contrib.gpt_assistant_agent import GPTAssistantAgent
-from prediction_market_agent_tooling.markets.data_models import AgentMarket
+from prediction_market_agent_tooling.markets.agent_market import AgentMarket
 
 from prediction_market_agent import utils
 from prediction_market_agent.agents.abstract import AbstractAgent

prediction_market_agent/agents/crewai_agent.py

@@ -1,7 +1,7 @@
 import json
 
 from langchain_community.tools import DuckDuckGoSearchRun
-from prediction_market_agent_tooling.markets.data_models import AgentMarket
+from prediction_market_agent_tooling.markets.agent_market import AgentMarket
 
 from prediction_market_agent.agents.abstract import AbstractAgent
 

prediction_market_agent/agents/custom_agent.py

@@ -2,7 +2,7 @@
 from typing import Optional
 
 import requests
-from prediction_market_agent_tooling.markets.data_models import AgentMarket
+from prediction_market_agent_tooling.markets.agent_market import AgentMarket
 from prediction_market_agent_tooling.tools.utils import (
     check_not_none,
     should_not_happen,
@@ -118,15 +118,17 @@ def answer_binary_market(self, market: AgentMarket) -> bool:
             elif tool_name := completion_dict.get("tool_name"):
                 tool_params = completion_dict.get("tool_params")
                 self.verbose_print(f"Agent: Using {tool_name=} with {tool_params=}.")
-                tool_output = (
+                tool_output = (  # type: ignore # Untyped for the sake of simplicity when the LLM is used.
                     self.google_search
                     if tool_name == "GoogleSearchTool"
                     else (
                         self.web_scrap_and_summarize
                         if tool_name == "WebScrapingTool"
                         else should_not_happen("Unknown tool requested from the LLM.")
                     )
-                )(**tool_params)
+                )(
+                    **tool_params
+                )
                 self.verbose_print(f"Tool: {tool_name=} returns {tool_output=}.")
                 messages.append(
                     Message(

prediction_market_agent/agents/langchain_agent.py

@@ -1,6 +1,6 @@
 from langchain.agents import AgentType, initialize_agent, load_tools
 from langchain_community.llms import OpenAI
-from prediction_market_agent_tooling.markets.data_models import AgentMarket
+from prediction_market_agent_tooling.markets.agent_market import AgentMarket
 
 from prediction_market_agent import utils
 from prediction_market_agent.agents.abstract import AbstractAgent
@@ -13,7 +13,9 @@ def __init__(self) -> None:
         # Can use pre-defined search tool
         # TODO: Tavily tool could give better results
         # https://docs.tavily.com/docs/tavily-api/langchain
-        tools = load_tools(["serpapi", "llm-math"], llm=llm, serpapi_api_key=keys.serp)
+        tools = load_tools(
+            ["serpapi", "llm-math"], llm=llm, serpapi_api_key=keys.serp_api_key
+        )
         self._agent = initialize_agent(
             tools, llm, agent=AgentType.ZERO_SHOT_REACT_DESCRIPTION, verbose=True
         )

prediction_market_agent/agents/llamaindex_agent.py

@@ -1,7 +1,7 @@
 from llama_index.agent import OpenAIAgent
 from llama_index.llms import OpenAI
 from llama_index.tools import FunctionTool
-from prediction_market_agent_tooling.markets.data_models import AgentMarket
+from prediction_market_agent_tooling.markets.agent_market import AgentMarket
 
 from prediction_market_agent import utils
 from prediction_market_agent.agents.abstract import AbstractAgent

prediction_market_agent/agents/metagpt_agent.py

@@ -3,7 +3,7 @@
 import os
 
 import dotenv
-from prediction_market_agent_tooling.markets.data_models import AgentMarket
+from prediction_market_agent_tooling.markets.agent_market import AgentMarket
 from prediction_market_agent_tooling.tools.utils import check_not_none
 
 from prediction_market_agent.agents.abstract import AbstractAgent

prediction_market_agent/tools/betting_strategies.py

@@ -6,8 +6,9 @@
 from prediction_market_agent_tooling.gtypes import Probability, wei_type, xDai
 from prediction_market_agent_tooling.markets.data_models import BetAmount, Currency
 from prediction_market_agent_tooling.markets.markets import MarketType
-from prediction_market_agent_tooling.markets.omen import (
-    OmenMarket,
+from prediction_market_agent_tooling.markets.omen.data_models import OmenMarket
+from prediction_market_agent_tooling.markets.omen.omen import (
+    OmenAgentMarket,
     omen_calculate_buy_amount,
 )
 from prediction_market_agent_tooling.tools.gnosis_rpc import GNOSIS_RPC_URL
@@ -57,6 +58,7 @@ def get_market_moving_bet(
     new_product - fixed_product = dx * na_y
     dx = (new_product - fixed_product) / na_y
     """
+    market_agent = OmenAgentMarket.from_data_model(market)
     amounts = market.outcomeTokenAmounts
     prices = check_not_none(
         market.outcomeTokenProbabilities, "No probabilities, is marked closed?"
@@ -98,7 +100,7 @@ def get_market_moving_bet(
         if check_vs_contract:
             expected_trade = omen_calculate_buy_amount(
                 web3=Web3(Web3.HTTPProvider(GNOSIS_RPC_URL)),
-                market=market,
+                market=market_agent,
                 investment_amount=wei_type(bet_amount),
                 outcome_index=bet_outcome_index,
             )
@@ -114,7 +116,7 @@ def get_market_moving_bet(
         if verbose:
             outcome = market.get_outcome_str(bet_outcome_index)
             print(
-                f"Target p_yes: {target_p_yes:.2f}, bet: {wei_to_xdai(bet_amount_wei):.2f}{market.BET_AMOUNT_CURRENCY} for {outcome}, new p_yes: {new_p_yes:.2f}"
+                f"Target p_yes: {target_p_yes:.2f}, bet: {wei_to_xdai(bet_amount_wei):.2f}{market_agent.currency} for {outcome}, new p_yes: {new_p_yes:.2f}"
             )
         if abs(target_p_yes - new_p_yes) < 0.01:
             break

prediction_market_agent/tools/google_search.py

@@ -4,7 +4,7 @@
 
 
 def google_search(query: str) -> list[str]:
-    params = {"q": query, "api_key": utils.APIKeys().serp, "num": 4}
+    params = {"q": query, "api_key": utils.APIKeys().serp_api_key, "num": 4}
     search = serpapi.GoogleSearch(params)
     urls = [result["link"] for result in search.get_dict()["organic_results"]]
     return urls

prediction_market_agent/utils.py

@@ -13,13 +13,13 @@ class APIKeys(APIKeysBase):
 
     @property
     def serp_api_key(self) -> str:
-        return check_not_none(  # type: ignore  # Remove once PMAT is correctly released and this doesn't ignore his typing.
+        return check_not_none(
             self.SERP_API_KEY, "SERP_API_KEY missing in the environment."
         )
 
     @property
     def openai_api_key(self) -> str:
-        return check_not_none(  # type: ignore  # Remove once PMAT is correctly released and this doesn't ignore his typing.
+        return check_not_none(
             self.OPENAI_API_KEY, "OPENAI_API_KEY missing in the environment."
         )
 
@@ -34,7 +34,7 @@ def get_market_prompt(question: str) -> str:
 
 
 def parse_result_to_boolean(result: str) -> bool:
-    return (  # type: ignore  # Remove once PMAT is correctly released and this doesn't ignore his typing.
+    return (
         True
         if result.lower() == "yes"
         else (

pyproject.toml

@@ -6,7 +6,7 @@ authors = ["Gnosis"]
 readme = "README.md"
 
 [tool.poetry.dependencies]
-python = ">=3.9,<3.10"
+python = "~3.10.0"
 python-dotenv = "*"
 openai = "*"
 langchain = "~0.1.1"
@@ -35,8 +35,7 @@ poetry = "^1.7.1"
 poetry-plugin-export = "^1.6.0"
 functions-framework = "^3.5.0"
 cron-validator = "^1.0.8"
-# TODO: Change to the correct version once it's released.
-prediction-market-agent-tooling = { git = "https://github.com/gnosis/prediction-market-agent-tooling.git", branch = "peter/fix-apikeys" }
+prediction-market-agent-tooling = "0.2.6"
 pydantic-settings = "^2.1.0"
 autoflake = "^2.2.1"
 isort = "^5.13.2"

tests/tools/test_betting_strategies.py

@@ -11,7 +11,7 @@
     xDai,
     xdai_type,
 )
-from prediction_market_agent_tooling.markets.markets import omen
+from prediction_market_agent_tooling.markets.omen import omen
 
 from prediction_market_agent.tools.betting_strategies import (
     get_kelly_criterion_bet,
@@ -59,7 +59,7 @@ def test_get_market_moving_bet_0(
         float(expected_buying_xdai_amount),
         atol=2.0,  # We don't expect it to be 100% accurate, but close enough.
     ), f"To move this martket to ~{wanted_p_yes_on_the_market}% for yes, the amount should be {expected_buying_xdai_amount}xDai, according to aiomen website."
-    assert outcome_index == TEST_OMEN_MARKET.get_outcome_index(
+    assert outcome_index == TEST_OMEN_MARKET.outcomes.index(
         expected_buying_outcome
     ), f"The buying outcome index should `{expected_buying_outcome}`."
 
@@ -79,4 +79,4 @@ def test_kelly_criterion_bet_0(est_p_yes: Probability, expected_outcome: str) ->
     )
     # Kelly estimates the best bet for maximizing the expected value of the logarithm of the wealth.
     # We don't know the real best xdai_amount, but at least we know which outcome index makes sense.
-    assert outcome_index == TEST_OMEN_MARKET.get_outcome_index(expected_outcome)
+    assert outcome_index == TEST_OMEN_MARKET.outcomes.index(expected_outcome)