[python] Add new agent type python-service

[eolivelli]

Summary:

• add a new component type "SERVICE"
• a Service doesn't use streaming to interact with external services
• the service is supposed to open some port on the pod and serve an HTTP service (no need to support TLS at the moment)
• initially this port is fixed and it must be 8000, in the future we can make it configurable
• add a new agent "python-service" that allows you to deploy a python application without consuming from a topic or producing messages
• add LangServe, FastAPI and Uvicorn to the bundled Python libraries

when you are running in docker the auto-reload feature works out of the box, as soon as you change your Python files the Python daemon is restarted

It is expected that the langstream gateway will act as a proxy to the service, adding:

• TLS support for public networks
• Authentication

Running Python application on Kubernetes

LangStream agents are already deployed as a StatefulSet (with persistent disks) and an headless Service. Deploying an application here allows you to easily leverage all the Kubernetes features, for instance:

• resource management (you can set set size of the machines, CPU, memory...)
• scalability (you can set the number of replicats)
• region awareness
• ...

Bonus feature added by this PR:

When you declare a Python agent you can easily pass environment variables from secrets:

module: default
name: "Basic LangServe application"
pipeline:
  - name: "Start LangServe"
    id: agent1
    type: "python-service"
    configuration:
      className: example.ChatBotService
      environment:
        - key: "OPENAI_API_KEY"
          value: "${secrets.open-ai.access-key}"
        - key: "LANGCHAIN_TRACING_V2"
          value: "true"
        - key: "LANGCHAIN_ENDPOINT"
          value: "${ secrets.lang-smith.api-url }"
        - key: "LANGCHAIN_API_KEY"
          value: "${ secrets.lang-smith.api-key }"          

This PR adds a sample application about how to integrate also with LangSmith