Add an endpoint for service metadata and use it to configure the app

[JoshuaC215]

This ended up being a big update!

• Introduce schema.ServiceMetadata which describes available agents, models, and default values based on the configuration (e.g. available LLM API keys)
• Introduce /info endpoint on the service to return the current metadata
• Update the AgentClient to retrieve the metadata by default and store it at AgentClient.info for use
• Update the Streamlit app to use the agent_client.info to set the available agents and models in the app settings
• While I was at it, refactored the error handling in AgentClient to be more idiomatic / robust / specific, and updated the app to give cleaner errors when there are connectivity issues
• Updated many tests to work based on all the changes above, and added tests for new functionality
• Now, with the dynamic config I could write an e2e integration tests that runs the app and hits the deployed service across docker containers 🎉

Example output for GET http://{service}/info

{
  "agents": [
    {
      "key": "research-assistant",
      "description": "A research assistant for generating research papers."
    },
    {
      "key": "chatbot",
      "description": "A simple chatbot."
    }
  ],
  "models": [
    "gpt-4o-mini",
    "claude-3-haiku",
  ],
  "default_agent": "research-assistant",
  "default_model": "gpt-4o-mini"
}

[codecov]

Codecov Report

Attention: Patch coverage is 80.68182% with 34 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/core/settings.py	43.47%	13 Missing ⚠️
src/streamlit_app.py	70.00%	9 Missing ⚠️
src/client/client.py	92.50%	6 Missing ⚠️
src/run_client.py	0.00%	4 Missing ⚠️
src/agents/research_assistant.py	33.33%	2 Missing ⚠️

Files with missing lines	Coverage Δ
src/agents/__init__.py	100.00% <100.00%> (ø)
src/agents/agents.py	100.00% <100.00%> (ø)
src/client/__init__.py	100.00% <100.00%> (ø)
src/schema/__init__.py	100.00% <ø> (ø)
src/schema/schema.py	85.45% <100.00%> (+2.47%)	⬆️
src/service/service.py	89.65% <100.00%> (+0.66%)	⬆️
src/agents/research_assistant.py	54.05% <33.33%> (-0.12%)	⬇️
src/run_client.py	0.00% <0.00%> (ø)
src/client/client.py	88.19% <92.50%> (-0.92%)	⬇️
src/streamlit_app.py	70.65% <70.00%> (-2.29%)	⬇️
... and 1 more