[HN Gopher] Show HN: Evolving Agents Framework
___________________________________________________________________
Show HN: Evolving Agents Framework
Hey HN, I've been working on an open-source framework for creating
AI agents that evolve, communicate, and collaborate to solve
complex tasks. The Evolving Agents Framework allows agents to:
Reuse, evolve, or create new agents dynamically based on semantic
similarity Communicate and delegate tasks to other specialized
agents Continuously improve by learning from past executions Define
workflows in YAML, making it easy to orchestrate agent interactions
Search for relevant tools and agents using OpenAI embeddings
Support multiple AI frameworks (BeeAI, etc.) Current Status &
Roadmap This is still a draft and a proof of concept (POC). Right
now, I'm focused on validating it in real-world scenarios to refine
and improve it. Next week, I'm adding a new feature to make it
useful for distributed multi-agent systems. This will allow agents
to work across different environments, improving scalability and
coordination. Why? Most agent-based AI frameworks today require
manual orchestration. This project takes a different approach by
allowing agents to decide and adapt based on the task at hand.
Instead of always creating new agents, it determines if existing
ones can be reused or evolved. Example Use Case: Let's say you
need an invoice analysis agent. Instead of manually configuring
one, our framework: Checks if a similar agent exists (e.g., a
document analyzer) Decides whether to reuse, evolve, or create a
new agent Runs the best agent and returns the extracted information
Here's a simple example in Python: import asyncio from
evolving_agents.smart_library.smart_library import SmartLibrary
from evolving_agents.core.llm_service import LLMService from
evolving_agents.core.system_agent import SystemAgent async def
main(): library = SmartLibrary("agent_library.json") llm =
LLMService(provider="openai", model="gpt-4o") system =
SystemAgent(library, llm) result = await
system.decide_and_act( request="I need an agent that
can analyze invoices and extract the total amount",
domain="document_processing", record_type="AGENT"
) print(f"Decision: {result['action']}") # 'reuse',
'evolve', or 'create' print(f"Agent:
{result['record']['name']}") if __name__ == "__main__":
asyncio.run(main()) Next Steps Validating in real-world use cases
and improving agent evolution strategies Adding distributed multi-
agent support for better scalability Full integration with BeeAI
Agent Communication Protocol (ACP) Better visualization tools for
debugging Would love feedback from the HN community! What features
would you like to see? Repo:
https://github.com/matiasmolinas/evolving-agents
Author : matiasmolinas
Score : 84 points
Date : 2025-03-09 16:51 UTC (5 hours ago)
(HTM) web link (github.com)
(TXT) w3m dump (github.com)
| matiasmolinas wrote:
| Here is the March 2025 BeeAI Community Call (
| https://www.youtube.com/watch?v=5-xqQBv-ccY ), where I presented
| the draft of the framework and shared some ideas on why it makes
| sense for me to provide this kind of framework and tools to AI
| agents.
| thierrydamiba wrote:
| Cool project! Thanks for sharing.
|
| Intrigued by the storage-why not use a VDB?
| EGreg wrote:
| I love the approach
|
| Would you like to join forces? Reach out to me:
| https://engageusers.ai/ecosystem.pdf
| esafak wrote:
| It's not continuously evolving outside the developer's control,
| right? Companies need to checkpoint dependencies for resilience.
| Can you define tests to ensure compliance of new versions?
| hleszek wrote:
| As long as we don't ask it to make paperclips it should be
| fine.
| unification_fan wrote:
| Is this distributed and peer to peer? Because I don't want to pay
| the cost of 50 agents myself. And we truly need something that
| can't be centralized
| khaledh wrote:
| There should be an image of Agent Smith somewhere in the readme
| (j/k) :)
| thrance wrote:
| Sure would beat the cringe ugly AI image currently in the
| readme.
| jfkimmes wrote:
| And here I was, thinking I was clever for coming up with the
| agent smith image for an agent framework.
|
| https://codeberg.org/jfkimmes/TinyAgentSmith
| kgwgk wrote:
| You're not alone: https://synw.github.io/agent-smith/
| antonkar wrote:
| Sadly we'll have many agent Smiths but as in the movies,
| without some rudimentary Matrix (it's possible to make some
| good democratic one), Neo and we will have no chance against
| them.
|
| Agents can change our world and us, but we have limited or no
| capability to change the online world and their often private
| multimodal "brains". We can represent them as familiar 3D
| environments and this way level the playing field.
|
| Agents are like strict librarians who spit quotes but we cannot
| enter the library. It has too many stolen things and companies
| don't want us to see that, they won't make 3D game-like
| representations of those libraries
| visarga wrote:
| Feels like the JavaScript framework wars era all over again.
| Wondering what will be the prevailing paradigm.
| Garlef wrote:
| No matter what paradigm will come out on top:
|
| Someone will notice there's something to be gained by building
| abstractions around caching/memoization and that running
| workflows in distributed environments might be augmented by
| some form of precompilation and/or hydration.
|
| Then others will boast that they can do everything with
| whatever flavour of vanilla they are familiar with.
|
| Then someone will bring up Greenspun's tenth rule.
|
| Anyways: We'll have pages and pages of HN threads to scroll
| through.
| kjhughes wrote:
| Your framework name suggests that you have an effective method of
| taking an existing agent that is "close" to meeting requirements
| by some similarity metric and evolving a new agent that will be
| better suited than the base agent is to meet the requirements.
|
| If this is true, your post, your repo README file, and your BeeAI
| Community call presentation (which starts here:
| https://www.youtube.com/watch?v=5-xqQBv-ccY&t=1294s) ought to be
| proclaiming such a notable success. Yet, I've seen little to
| nothing about it in any of those places. Am I missing it?
|
| How exactly does your agent evolution process work?
| bob1029 wrote:
| It appears to be a "please improve yourself" style prompt that
| is ran through the LLM.
|
| https://github.com/matiasmolinas/evolving-agents?tab=readme-...
| Splizard wrote:
| > The second part demonstrates how agents communicate with each
| other through workflows defined in YAML
|
| oh no
___________________________________________________________________
(page generated 2025-03-09 22:00 UTC)