GenAgent Utilities
A Python package providing utilities for generative agent tasks, including:
- Interacting with LLM providers (OpenAI, Anthropic).
- Helper functions for prompt engineering and structured output.
- Basic agent memory and chat session management.
(See genagent/__init__.py for the current version)
Installation
pip install genagent
Configuration
GenAgent now supports environment variable configuration for default provider and model:
# In your .env file or environment
DEFAULT_PROVIDER=ant # or 'oai' for OpenAI
DEFAULT_MODEL=claude-3-7-sonnet-latest # or any model you prefer
OPENAI_API_KEY=your_openai_key
ANTHROPIC_API_KEY=your_anthropic_key
Usage
from genagent import gen, Agent, ChatSession, mod_gen, modular_instructions
# --- Basic text generation ---
response = gen("Tell me a concise joke.")
print(f"Joke: {response}")
# --- Agent memory example ---
my_agent = Agent(name="MyTestAgent")
my_agent.add_memory("The capital of France is Paris.")
my_agent.add_memory("The Eiffel Tower is in Paris.")
retrieved_text = my_agent.retrieve_memories_as_text(query="What is a famous landmark in the capital of France?")
print(f"Retrieved memories based on query: {retrieved_text}")
# --- Chat session example ---
# Assuming 'my_agent' is already created and has some memories
chat_session = ChatSession(agent=my_agent, system_prompt="You are a helpful assistant. Use your memory when relevant.")
user_query = "What can you tell me about Paris based on your memory?"
assistant_response = chat_session.chat(user_query)
print(f"\nChat Session:")
print(f" User: {user_query}")
print(f" Assistant: {assistant_response}")
# --- Structured output with mod_gen ---
# 'mod_gen' helps generate structured output based on a list of instruction modules.
# Each module in the list is a dictionary that can have:
# - 'instruction': (Required) The text prompt or instruction for the LLM.
# - 'name': (Optional) If provided, the LLM's output for this instruction will be captured under this key in the result.
# Modules without a 'name' contribute to the context but aren't extracted as separate fields.
# Example modules for extracting information:
instruction_modules = [
{
"instruction": "You are analyzing a piece of text. The text is: 'GenAgent is a Python library for building applications with Large Language Models. It simplifies interactions with models like GPT and Claude, and provides tools for memory management.'"
},
{
"name": "summary",
"instruction": "Provide a brief one-sentence summary of the text."
},
{
"name": "main_features",
"instruction": "List up to three main features or capabilities mentioned."
},
{
"name": "target_audience",
"instruction": "Based on the description, who is the target audience?"
}
]
print("\nAttempting structured output with mod_gen...")
# mod_gen now uses environment defaults if not specified
structured_output = mod_gen(
modules=instruction_modules
# provider and model will use environment defaults
)
if structured_output:
print("\nStructured Output from mod_gen:")
for key, value in structured_output.items():
print(f"- {key.capitalize()}: {value}")
else:
print("\nMod_gen did not return structured output. Check for errors or try debugging.")
What's New in 0.2.1
- Parallelization: Added experiment utils
run_parallel_dict_mapandrun_parallel_dict_productfrom e chi
What's New in 0.2.0
- Modular architecture: Agent utilities moved to separate
agent_utilsmodule - Environment configuration: Default provider and model can be set via environment variables
- Improved defaults:
mod_gennow uses configured defaults instead of hardcoded values - Better code organization: Cleaner separation of concerns between LLM utilities and agent functionality
Development
To install for development (from the root of the llm-utils repository):
pip install -e .
License
This project is licensed under the MIT License - see the LICENSE file for details.
Metadata
Release files for genagent 0.4.1
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| genagent-0.4.1.tar.gz | 19.7 kB | Without provenance | Details for genagent-0.4.1.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| genagent-0.4.1-py3-none-any.whl | Python 3 | none | any | Details for genagent-0.4.1-py3-none-any.whl |
Total release size: 38.3 kB
Release files / genagent-0.4.1.tar.gz
| Download URL | genagent-0.4.1.tar.gz |
|---|---|
| Size | 19.7 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
5b3c75dc028ea71654cee4a22e65564bc956862f069c2f7030667d758372f273
|
|
BLAKE2b-256 checksum How to use checksums |
29bec618cc67a6efc1891696bdc72c3dcc31740a7445b18229194f447e08799f
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.11.12
|
Release files / genagent-0.4.1-py3-none-any.whl
| Download URL | genagent-0.4.1-py3-none-any.whl |
|---|---|
| Size | 18.7 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
caf7bbc1c6308bfdf14f6560a10e89a3c34323e42413df751481815194802ac6
|
|
BLAKE2b-256 checksum How to use checksums |
f1ff50d2a0e9b96be16eddc65ac1ed450ae067110c4953d33551d7525139af1a
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.11.12
|
Security for genagent 0.4.1
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.