Introducing AutoBuilder

[LinxinS97]

Why are these changes needed?

AutoBuild can help users build a multi-agent group chat for a specific task more easily. Here is a sample case:

from autogen.agentchat.contrib.agent_builder import AgentBuilder

# Specify a building task with a general description. 
# Building task will help build manager (a LLM) decide what agents should be build.
building_task = "Find a paper on arxiv by programming, and analysis its application in some domain. For example, find a latest paper about gpt-4 on arxiv and find its potential applications in software."

# Initialize a builder
builder = AgentBuilder(config_path=config_path, builder_model='gpt-4-1106-preview', agent_model='gpt-4-1106-preview')

# Let the builder generate the group chat agents' configs, including agent name and system message.
builder.build(building_task, default_llm_config)

# Let agents generated in build() complete the task collaboratively in a group chat.
builder.start(task="Find a latest paper about gpt-4 on arxiv and find its potential applications in software.")

# Clear agent caches
builder.clear_all_agents(recycle_endpoint=True)

Related issue number

Checks

• I've included any doc changes needed for https://microsoft.github.io/autogen/. See https://microsoft.github.io/autogen/docs/Contribute#documentation to build and test documentation locally.
• I've added tests (if relevant) corresponding to the changes introduced in this PR.
• Blog
• I've made sure all auto checks have passed.

[LinxinS97]

This PR has been merged by PR #846 #855 #856