PyPI | Documentation | Discord

Language InterOperable Network - LION

lionagi version 0.2.0 nightly - alpha

Powerful Intelligent Workflow Automation

lionagi is an intelligent agentic workflow automation framework. It introduces advanced ML models into any existing workflows and data infrastructure.

Why Automating Workflows?

Intelligent AI models such as Large Language Model (LLM), introduced new possibilities of human-computer interaction. LLMs is drawing a lot of attention worldwide due to its “one model fits all”, and incredible performance. One way of using LLM is to use as search engine, however, this usage is complicated by the fact that LLMs hallucinate.

What goes inside of a LLM is more akin to a black-box, lacking interpretability, meaning we don’t know how it reaches certain answer or conclusion, thus we cannot fully trust/rely the output from such a system. Another approach of using LLM is to treat them as intelligent agent, that are equipped with various tools and data sources. A workflow conducted by such an intelligent agent have clear steps, and we can specify, observe, evaluate and optimize the logic for each decision that the agent made to perform actions. This approach, though we still cannot pinpoint how LLM output what it outputs, but the flow itself is explainable.

Version Checklist

✅ : Done

〰️ : Not started

🛠️: In progress

Folder	Module	written	Can run	Doc
Action	function_calling	✅	✅	〰️
	manual	〰️	〰️	〰️
	node	✅	✅	〰️
	tool_manager	✅	✅	〰️
	tool	✅	✅	〰️
agent	base_agent	✅	🛠️	〰️
agent.evaluator	evaluator	🛠️	🛠️	〰️
	score	🛠️	🛠️	〰️
	vote	🛠️	🛠️	〰️
agent.learner	learn	🛠️	🛠️	〰️
	learner	🛠️	🛠️	〰️
agent.planner	plan	🛠️	🛠️	〰️
collections.abc	component	✅	✅	〰️
	concepts	✅	✅	〰️
	exceptions	✅	✅	〰️
collections	exchange	✅	✅	〰️
	flow	✅	✅	〰️
	model	✅	✅	〰️
	pile	✅	✅	〰️
	progression	✅	✅	〰️
directive.engine	ast_evaluator	🛠️	〰️	〰️
	sandbox	🛠️	〰️	〰️
	script_engine	🛠️	〰️	〰️
directive.parser	base	🛠️	〰️	〰️
	syntax.txt	🛠️	〰️	〰️
directive.template	base	🛠️	〰️	〰️
	schema	🛠️	〰️	〰️
directive.unit	templates	✅	✅	〰️
	unit	✅	🛠️	〰️
	parallel_unit	🛠️	🛠️	〰️
director	direct	🛠️	〰️	〰️
	director	🛠️	〰️	〰️
execute	base	🛠️	〰️	〰️
	branch_executor	🛠️	〰️	〰️
	instruction_map	🛠️	〰️	〰️
	neo4j_executor	🛠️	〰️	〰️
	structure_executor	🛠️	〰️	〰️
generic	edge	✅	✅	〰️
	graph	✅	✅	〰️
	node	✅	✅	〰️
	hyperedge	🛠️		〰️
	tree_node	✅	✅	〰️
	tree	✅	✅	〰️
mail	mail	✅	✅	〰️
	mail_manager	✅	✅	〰️
	package	✅	✅	〰️
	start_mail	✅	✅	〰️
message	action_request	✅	✅	〰️
	action_response	✅	✅	〰️
	assistant_response	✅	✅	〰️
	instruction	✅	✅	〰️
	message	✅	✅	〰️
	system	✅	✅	〰️
report	base	✅	✅	〰️
	form	✅	✅	〰️
	report	✅	✅	〰️
rule	base	✅	✅	〰️
	action	✅	✅	〰️
	boolean	✅	✅	〰️
	choice	✅	✅	〰️
	mapping	✅	✅	〰️
	number	✅	✅	〰️
	rulebook	✅	✅	〰️
	string	✅	✅	〰️
session	branch	✅	✅	〰️
	directive-mixin	🛠️	〰️	〰️
	session	🛠️	〰️	〰️
validator	validator	✅	✅	〰️
structure	chain	🛠️	🛠️	〰️
	tree	〰️	〰️	〰️
	graph	〰️	〰️	〰️
	forest	〰️	〰️	〰️
work	work function	✅	✅	〰️
	work queue	✅	✅	〰️
	work	✅	✅	〰️
	worker	✅	✅	〰️
	worklog	✅	✅	〰️

Community

We encourage contributions to LionAGI and invite you to enrich its features and capabilities. Engage with us and other community members Join Our Discord

Citation

When referencing LionAGI in your projects or research, please cite:

@software{Li_LionAGI_2023,
  author = {Haiyang Li},
  month = {12},
  year = {2023},
  title = {LionAGI: Towards Automated General Intelligence},
  url = {https://github.com/lion-agi/lionagi},
}

Requirements

Python 3.10 or higher.