GenAgent

0.4.1 verified Fri May 01 auth: no python

A Python library providing utilities for generative agent tasks, including LLM interactions and agent memory. Current version 0.4.1, requires Python >=3.8. Development is active with recent updates.

pip install genagent

Common errors

error ModuleNotFoundError: No module named 'genagent.agent' ↓

cause Importing from submodule 'genagent.agent' which was removed in 0.4.0.

fix

Change import to 'from genagent import Agent'

error AttributeError: 'Memory' object has no attribute 'save' ↓

cause Trying to save memory manually, but Memory does not have a save method; use persistence_file at initialization.

fix

Initialize Memory with persistence_file='/path/to/memory.json' and it will auto-save.

error TypeError: __init__() got an unexpected keyword argument 'api_key' ↓

cause Incorrect initialization of LLMClient; the parameter name might have changed in newer versions.

fix

Check the latest docs: use 'api_token' or 'auth_token' depending on version. In 0.4.1, use 'api_key' as shown.

Warnings

breaking In version 0.4.0, the 'Agent' class was moved from genagent.agent to the top-level package. Code using 'from genagent.agent import Agent' will break. ↓

fix Use 'from genagent import Agent' instead.

gotcha The Memory class is not automatically persisted. If you need long-term memory, you must implement your own persistence layer or use the optional persistence feature with a file path. ↓

fix Pass a persistence file: Memory(persistence_file='/path/to/memory.json')

deprecated The 'genagent.tools' module is deprecated since 0.4.0 and will be removed in a future release. Use the top-level 'genagent.tools' as a package instead. ↓

fix Migrate to using 'from genagent import tools'.

Imports

Agent
wrong
```
from genagent.agent import Agent
```
correct
```
from genagent import Agent
```
Agent is exported from the top-level package, not a submodule.
LLMClient
```
from genagent import LLMClient
```
LLMClient is a top-level export.

Memory

wrong

from genagent.memory import Memory

correct

from genagent import Memory

Memory is exported from the top-level package.

Quickstart

Initialize a generative agent with an LLM client and memory, then run a query.

from genagent import Agent, LLMClient, Memory

memory = Memory()
client = LLMClient(api_key=os.environ.get('OPENAI_API_KEY', ''))
agent = Agent(client=client, memory=memory)
response = agent.run("What is the capital of France?")
print(response)