prediction-market-agent-tooling 0.29.0__tar.gz → 0.32.0__tar.gz

{prediction_market_agent_tooling-0.29.0 → prediction_market_agent_tooling-0.32.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: prediction-market-agent-tooling
-Version: 0.29.0
+Version: 0.32.0
 Summary: Tools to benchmark, deploy and monitor prediction market agents.
 Author: Gnosis
 Requires-Python: >=3.10,<3.12
@@ -61,7 +61,6 @@ Deploying and monitoring agents using GCP requires that you set up the gcloud CL
 
 ```bash
 MANIFOLD_API_KEY=...
-BET_FROM_ADDRESS=...
 BET_FROM_PRIVATE_KEY=...
 OPENAI_API_KEY=...
 ```
@@ -76,46 +75,54 @@ For example:
 
 ```python
 import prediction_market_agent_tooling.benchmark.benchmark as bm
+from prediction_market_agent_tooling.benchmark.agents import RandomAgent
 from prediction_market_agent_tooling.markets.markets import MarketType, get_binary_markets
 
 benchmarker = bm.Benchmarker(
     markets=get_binary_markets(limit=10, market_type=MarketType.MANIFOLD),
-    agents=[...],
+    agents=[RandomAgent(agent_name="a_random_agent")],
 )
 benchmarker.run_agents()
 md = benchmarker.generate_markdown_report()
 ```
 
-This produces a markdown report comparing agents:
+This produces a markdown report that you can use for comparing agents side-by-side, like:
 
 ![Benchmark results](assets/comparison-report.png)
 
 ## Deploying
 
-Create a deployable agent by subclassing the `DeployableAgent` base class, and implementing the 
+> **Deprecated**: We suggest using your own infrastructure to deploy, but you may still find this useful.
+
+Create a deployable agent by subclassing the `DeployableTraderAgent` base class, and implementing the `answer_binary_market` method.
 
 For example, deploy an agent that randomly picks an outcome:
 
 ```python
 import random
-from prediction_market_agent_tooling.deploy.agent import DeployableAgent
+from prediction_market_agent_tooling.deploy.agent import DeployableTraderAgent
 from prediction_market_agent_tooling.markets.agent_market import AgentMarket
 
-class DeployableCoinFlipAgent(DeployableAgent):
+class DeployableCoinFlipAgent(DeployableTraderAgent):
     def answer_binary_market(self, market: AgentMarket) -> bool | None:
         return random.choice([True, False])
 
 DeployableCoinFlipAgent().deploy_gcp(...)
 ```
 
-For deploying a Safe manually for a given agent, run the script below:
+### Safe
+
+Agents can control funds via a wallet primary key only, or optionally via a [Safe](https://safe.global/) as well. For deploying a Safe manually for a given agent, run the script below:
 
 ```commandline
 poetry run python scripts/create_safe_for_agent.py  --from-private-key <YOUR_AGENT_PRIVATE_KEY> --salt-nonce 42
 ```
+
 This will output the newly created Safe in the terminal, and it can then be copied over to the deployment part (e.g. Terraform).
 Note that `salt_nonce` can be passed so that the created safe is deterministically created for each agent, so that, if the same `salt_nonce` is used, the script will not create a new Safe for the agent, instead it will output the previously existent Safe.
 
+You can then specify this agent's Safe address with the `SAFE_ADDRESS` environment variable.
+
 ## Monitoring
 
 Monitor the performance of the agents deployed to GCP, as well as meta-metrics of the prediction market platforms they are deployed to.
@@ -132,15 +139,42 @@ Which launches in the browser:
 
 ## The Market Platforms
 
-The following markets platforms are supported:
+The following prediction market platforms are supported:
+
+| Platform                              | Benchmarking | Deployment | Monitoring |
+|---------------------------------------|--------------|------------|------------|
+| [Manifold](https://manifold.markets/) | ✅ | ✅ | ✅ |
+| [AIOmen](https://aiomen.eth.limo/)    | ✅ | ✅ | ✅ |
+| [Polymarket](https://polymarket.com/) | ✅ | ❌ | ❌ |
+
+## Prediction Markets Python API
+
+We have built clean abstractions for taking actions on the different prediction market platforms (retrieving markets, buying and selling tokens, etc.). This is currently undocumented, but for now, inspecting the [`AgentMarket`](https://github.com/gnosis/prediction-market-agent-tooling/blob/1e497fff9f2b53e4e3e1beb5dda08b4d49da881b/prediction_market_agent_tooling/markets/agent_market.py) class and its methods is your best bet.
+
+For example:
+
+```python
+from prediction_market_agent_tooling.config import APIKeys
+from prediction_market_agent_tooling.markets.agent_market import SortBy
+from prediction_market_agent_tooling.markets.omen.omen import OmenAgentMarket
+
+# Place a bet on the market closing soonest
+market = OmenAgentMarket.get_binary_markets(limit=1, sort_by=SortBy.CLOSING_SOONEST)[0]
+market.place_bet(outcome=True, amount=market.get_bet_amount(0.1))
+
+# View your positions
+my_positions = OmenAgentMarket.get_positions(user_id=APIKeys().bet_from_address)
+print(my_positions)
+
+# Sell position (accounting for fees)
+market.sell_tokens(outcome=True, amount=market.get_bet_amount(0.095))
+```
 
-- [Manifold](https://manifold.markets/)
-- [AIOmen](https://aiomen.eth.limo/)
-- [Polymarket](https://polymarket.com/) - Benchmarking only. Deploy and monitor TODO
+This API can be built on top of to create your application. See [here](https://github.com/gnosis/prediction-market-agent/tree/main/prediction_market_agent/agents/microchain_agent) for an example.
 
 ## Contributing
 
-See the [Issues](https://github.com/gnosis/prediction-market-agent-tooling/issues) for ideas of things that need fixing or implementing. Or come up with your own :D.
+See the [Issues](https://github.com/gnosis/prediction-market-agent-tooling/issues) for ideas of things that need fixing or implementing. The team is also receptive to new issues and PRs.
 
 We use `mypy` for static type checking, and `isort`, `black` and `autoflake` for linting. These all run as steps in CI.
 

{prediction_market_agent_tooling-0.29.0 → prediction_market_agent_tooling-0.32.0}/README.md

@@ -18,7 +18,6 @@ Deploying and monitoring agents using GCP requires that you set up the gcloud CL
 
 ```bash
 MANIFOLD_API_KEY=...
-BET_FROM_ADDRESS=...
 BET_FROM_PRIVATE_KEY=...
 OPENAI_API_KEY=...
 ```
@@ -33,46 +32,54 @@ For example:
 
 ```python
 import prediction_market_agent_tooling.benchmark.benchmark as bm
+from prediction_market_agent_tooling.benchmark.agents import RandomAgent
 from prediction_market_agent_tooling.markets.markets import MarketType, get_binary_markets
 
 benchmarker = bm.Benchmarker(
     markets=get_binary_markets(limit=10, market_type=MarketType.MANIFOLD),
-    agents=[...],
+    agents=[RandomAgent(agent_name="a_random_agent")],
 )
 benchmarker.run_agents()
 md = benchmarker.generate_markdown_report()
 ```
 
-This produces a markdown report comparing agents:
+This produces a markdown report that you can use for comparing agents side-by-side, like:
 
 ![Benchmark results](assets/comparison-report.png)
 
 ## Deploying
 
-Create a deployable agent by subclassing the `DeployableAgent` base class, and implementing the 
+> **Deprecated**: We suggest using your own infrastructure to deploy, but you may still find this useful.
+
+Create a deployable agent by subclassing the `DeployableTraderAgent` base class, and implementing the `answer_binary_market` method.
 
 For example, deploy an agent that randomly picks an outcome:
 
 ```python
 import random
-from prediction_market_agent_tooling.deploy.agent import DeployableAgent
+from prediction_market_agent_tooling.deploy.agent import DeployableTraderAgent
 from prediction_market_agent_tooling.markets.agent_market import AgentMarket
 
-class DeployableCoinFlipAgent(DeployableAgent):
+class DeployableCoinFlipAgent(DeployableTraderAgent):
     def answer_binary_market(self, market: AgentMarket) -> bool | None:
         return random.choice([True, False])
 
 DeployableCoinFlipAgent().deploy_gcp(...)
 ```
 
-For deploying a Safe manually for a given agent, run the script below:
+### Safe
+
+Agents can control funds via a wallet primary key only, or optionally via a [Safe](https://safe.global/) as well. For deploying a Safe manually for a given agent, run the script below:
 
 ```commandline
 poetry run python scripts/create_safe_for_agent.py  --from-private-key <YOUR_AGENT_PRIVATE_KEY> --salt-nonce 42
 ```
+
 This will output the newly created Safe in the terminal, and it can then be copied over to the deployment part (e.g. Terraform).
 Note that `salt_nonce` can be passed so that the created safe is deterministically created for each agent, so that, if the same `salt_nonce` is used, the script will not create a new Safe for the agent, instead it will output the previously existent Safe.
 
+You can then specify this agent's Safe address with the `SAFE_ADDRESS` environment variable.
+
 ## Monitoring
 
 Monitor the performance of the agents deployed to GCP, as well as meta-metrics of the prediction market platforms they are deployed to.
@@ -89,14 +96,41 @@ Which launches in the browser:
 
 ## The Market Platforms
 
-The following markets platforms are supported:
+The following prediction market platforms are supported:
+
+| Platform                              | Benchmarking | Deployment | Monitoring |
+|---------------------------------------|--------------|------------|------------|
+| [Manifold](https://manifold.markets/) | ✅ | ✅ | ✅ |
+| [AIOmen](https://aiomen.eth.limo/)    | ✅ | ✅ | ✅ |
+| [Polymarket](https://polymarket.com/) | ✅ | ❌ | ❌ |
+
+## Prediction Markets Python API
+
+We have built clean abstractions for taking actions on the different prediction market platforms (retrieving markets, buying and selling tokens, etc.). This is currently undocumented, but for now, inspecting the [`AgentMarket`](https://github.com/gnosis/prediction-market-agent-tooling/blob/1e497fff9f2b53e4e3e1beb5dda08b4d49da881b/prediction_market_agent_tooling/markets/agent_market.py) class and its methods is your best bet.
+
+For example:
+
+```python
+from prediction_market_agent_tooling.config import APIKeys
+from prediction_market_agent_tooling.markets.agent_market import SortBy
+from prediction_market_agent_tooling.markets.omen.omen import OmenAgentMarket
+
+# Place a bet on the market closing soonest
+market = OmenAgentMarket.get_binary_markets(limit=1, sort_by=SortBy.CLOSING_SOONEST)[0]
+market.place_bet(outcome=True, amount=market.get_bet_amount(0.1))
+
+# View your positions
+my_positions = OmenAgentMarket.get_positions(user_id=APIKeys().bet_from_address)
+print(my_positions)
+
+# Sell position (accounting for fees)
+market.sell_tokens(outcome=True, amount=market.get_bet_amount(0.095))
+```
 
-- [Manifold](https://manifold.markets/)
-- [AIOmen](https://aiomen.eth.limo/)
-- [Polymarket](https://polymarket.com/) - Benchmarking only. Deploy and monitor TODO
+This API can be built on top of to create your application. See [here](https://github.com/gnosis/prediction-market-agent/tree/main/prediction_market_agent/agents/microchain_agent) for an example.
 
 ## Contributing
 
-See the [Issues](https://github.com/gnosis/prediction-market-agent-tooling/issues) for ideas of things that need fixing or implementing. Or come up with your own :D.
+See the [Issues](https://github.com/gnosis/prediction-market-agent-tooling/issues) for ideas of things that need fixing or implementing. The team is also receptive to new issues and PRs.
 
 We use `mypy` for static type checking, and `isort`, `black` and `autoflake` for linting. These all run as steps in CI.

{prediction_market_agent_tooling-0.29.0 → prediction_market_agent_tooling-0.32.0}/prediction_market_agent_tooling/deploy/agent.py

@@ -273,11 +273,11 @@ class DeployableTraderAgent(DeployableAgent):
         for market in markets:
             result = self.answer_binary_market(market)
             if result is None:
-                logger.debug(f"Skipping market {market} as no answer was provided")
+                logger.info(f"Skipping market {market} as no answer was provided")
                 continue
             if self.place_bet:
                 amount = self.calculate_bet_amount(result, market)
-                logger.debug(
+                logger.info(
                     f"Placing bet on {market} with result {result} and amount {amount}"
                 )
                 market.place_bet(

{prediction_market_agent_tooling-0.29.0 → prediction_market_agent_tooling-0.32.0}/prediction_market_agent_tooling/loggers.py

@@ -14,12 +14,21 @@ class LogFormat(str, Enum):
     GCP = "gcp"
 
 
+class LogLevel(str, Enum):
+    CRITICAL = "CRITICAL"
+    ERROR = "ERROR"
+    WARNING = "WARNING"
+    INFO = "INFO"
+    DEBUG = "DEBUG"
+
+
 class LogConfig(BaseSettings):
     model_config = SettingsConfigDict(
         env_file=".env", env_file_encoding="utf-8", extra="ignore"
     )
 
     LOG_FORMAT: LogFormat = LogFormat.DEFAULT
+    LOG_LEVEL: LogLevel = LogLevel.DEBUG
 
 
 GCP_LOG_LOGURU_FORMAT = (
@@ -49,7 +58,7 @@ def patch_logger() -> None:
     # Change built-in logging.
     if format_logging is not None:
         logging.basicConfig(
-            level=logging.DEBUG, format=format_logging, datefmt=datefmt_logging
+            level=config.LOG_LEVEL.value, format=format_logging, datefmt=datefmt_logging
         )
 
     # Change loguru.
@@ -58,7 +67,7 @@ def patch_logger() -> None:
         logger.add(
             sys.stdout,
             format=format_loguru,
-            level="DEBUG",  # Can be the lowest level, higher ones will use by default this one.
+            level=config.LOG_LEVEL.value,
             colorize=True,
         )
 

{prediction_market_agent_tooling-0.29.0 → prediction_market_agent_tooling-0.32.0}/prediction_market_agent_tooling/markets/agent_market.py

@@ -5,6 +5,7 @@ from enum import Enum
 from eth_typing import ChecksumAddress
 from pydantic import BaseModel, field_validator
 
+from prediction_market_agent_tooling.config import APIKeys
 from prediction_market_agent_tooling.gtypes import Probability
 from prediction_market_agent_tooling.markets.data_models import (
     Bet,
@@ -196,3 +197,7 @@ class AgentMarket(BaseModel):
         if self.is_closed() or not self.has_liquidity():
             return False
         return True
+
+    @classmethod
+    def get_user_url(cls, keys: APIKeys) -> str:
+        raise NotImplementedError("Subclasses must implement this method")

{prediction_market_agent_tooling-0.29.0 → prediction_market_agent_tooling-0.32.0}/prediction_market_agent_tooling/markets/manifold/manifold.py

@@ -11,6 +11,7 @@ from prediction_market_agent_tooling.markets.agent_market import (
 )
 from prediction_market_agent_tooling.markets.data_models import BetAmount, Currency
 from prediction_market_agent_tooling.markets.manifold.api import (
+    get_authenticated_user,
     get_manifold_binary_markets,
     place_bet,
 )
@@ -108,3 +109,7 @@ class ManifoldAgentMarket(AgentMarket):
                 excluded_questions=excluded_questions,
             )
         ]
+
+    @classmethod
+    def get_user_url(cls, keys: APIKeys) -> str:
+        return get_authenticated_user(keys.manifold_api_key.get_secret_value()).url

{prediction_market_agent_tooling-0.29.0 → prediction_market_agent_tooling-0.32.0}/prediction_market_agent_tooling/markets/omen/omen.py

@@ -156,7 +156,7 @@ class OmenAgentMarket(AgentMarket):
         )
 
     def sell_tokens(
-        self, outcome: bool, amount: TokenAmount, auto_withdraw: bool = True
+        self, outcome: bool, amount: TokenAmount, auto_withdraw: bool = False
     ) -> None:
         if not self.can_be_traded():
             raise ValueError(
@@ -369,6 +369,10 @@ class OmenAgentMarket(AgentMarket):
 
         return positions
 
+    @classmethod
+    def get_user_url(cls, keys: APIKeys) -> str:
+        return f"https://gnosisscan.io/address/{keys.bet_from_address}"
+
 
 def pick_binary_market(
     sort_by: SortBy = SortBy.CLOSING_SOONEST, filter_by: FilterBy = FilterBy.OPEN

{prediction_market_agent_tooling-0.29.0 → prediction_market_agent_tooling-0.32.0}/prediction_market_agent_tooling/markets/omen/omen_subgraph_handler.py

@@ -205,8 +205,9 @@ class OmenSubgraphHandler(metaclass=SingletonMeta):
                 finalized_before
             )
 
+        # `excluded_question_titles` can not be an empty list, otherwise the API bugs out and returns nothing.
         excluded_question_titles = [""]
-        if excluded_questions is not None:
+        if excluded_questions:
             excluded_question_titles = [i for i in excluded_questions]
 
         where_stms["question_"]["title_not_in"] = excluded_question_titles
@@ -439,6 +440,8 @@ class OmenSubgraphHandler(metaclass=SingletonMeta):
         market_id: t.Optional[ChecksumAddress] = None,
         filter_by_answer_finalized_not_null: bool = False,
         type_: t.Literal["Buy", "Sell"] | None = None,
+        market_opening_after: datetime | None = None,
+        collateral_amount_more_than: Wei | None = None,
     ) -> list[OmenBet]:
         if not end_time:
             end_time = utcnow()
@@ -457,6 +460,12 @@ class OmenSubgraphHandler(metaclass=SingletonMeta):
             where_stms.append(trade.fpmm == market_id.lower())
         if filter_by_answer_finalized_not_null:
             where_stms.append(trade.fpmm.answerFinalizedTimestamp != None)
+        if market_opening_after is not None:
+            where_stms.append(
+                trade.fpmm.openingTimestamp > to_int_timestamp(market_opening_after)
+            )
+        if collateral_amount_more_than is not None:
+            where_stms.append(trade.collateralAmount > collateral_amount_more_than)
 
         trades = self.trades_subgraph.Query.fpmmTrades(
             first=sys.maxsize, where=where_stms
@@ -473,6 +482,8 @@ class OmenSubgraphHandler(metaclass=SingletonMeta):
         end_time: t.Optional[datetime] = None,
         market_id: t.Optional[ChecksumAddress] = None,
         filter_by_answer_finalized_not_null: bool = False,
+        market_opening_after: datetime | None = None,
+        collateral_amount_more_than: Wei | None = None,
     ) -> list[OmenBet]:
         return self.get_trades(
             better_address=better_address,
@@ -481,6 +492,8 @@ class OmenSubgraphHandler(metaclass=SingletonMeta):
             market_id=market_id,
             filter_by_answer_finalized_not_null=filter_by_answer_finalized_not_null,
             type_="Buy",  # We consider `bet` to be only the `Buy` trade types.
+            market_opening_after=market_opening_after,
+            collateral_amount_more_than=collateral_amount_more_than,
         )
 
     def get_resolved_bets(

{prediction_market_agent_tooling-0.29.0 → prediction_market_agent_tooling-0.32.0}/prediction_market_agent_tooling/tools/is_predictable.py

@@ -1,5 +1,6 @@
 from loguru import logger
 
+from prediction_market_agent_tooling.config import APIKeys
 from prediction_market_agent_tooling.tools.cache import persistent_inmemory_cache
 
 # I tried to make it return a JSON, but it didn't work well in combo with asking it to do chain of thought.
@@ -44,7 +45,11 @@ def is_predictable_binary(
         logger.info("langchain not installed, skipping is_predictable_binary")
         return True
 
-    llm = ChatOpenAI(model=engine, temperature=0.0)
+    llm = ChatOpenAI(
+        model=engine,
+        temperature=0.0,
+        api_key=APIKeys().openai_api_key.get_secret_value(),
+    )
 
     prompt = ChatPromptTemplate.from_template(template=prompt_template)
     messages = prompt.format_messages(question=question)

{prediction_market_agent_tooling-0.29.0 → prediction_market_agent_tooling-0.32.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "prediction-market-agent-tooling"
-version = "0.29.0"
+version = "0.32.0"
 description = "Tools to benchmark, deploy and monitor prediction market agents."
 authors = ["Gnosis"]
 readme = "README.md"