feat(langchain): Add support for LangSmith and executing RunnableConfig for our Runnables

[kaushikb11]

Overview

This PR introduces support for LangSmith and executing RunnableConfig for our Runnables, enhancing users' ability to trace, debug and monitor Guardrails and execute the Runnables with user defined configuration.

We have also introduced BaseRunnable to improve code reusability and maintain consistency across both Runnable implementations.

Guard Runnable

guard = Guard().use(CompetitorCheck(competitors=competitors_list, on_fail="fix"))
prompt = ChatPromptTemplate.from_template("Answer this question {question}")
output_parser = StrOutputParser()

chain = prompt | model | output_parser | guard.to_runnable()

result = chain.invoke({"question": "Tell me five Italian car companies?"})

Validator Runnable

regex_match = RegexMatch(
    "Ice cream", match_type="search", on_fail="refrain"
).to_runnable()
reading_time = ReadingTime(0.05, on_fail="refrain").to_runnable()
output_parser = StrOutputParser()

chain = prompt | model | regex_match | reading_time | output_parser
result = chain.invoke({"topic": topic})

[CalebCourier]

@kaushikb11 please address any checks that are failing. Thanks!

[kaushikb11]

@CalebCourier Fixed the tests. Could you please run (approve) the workflow?

[CalebCourier]

Since this is from a fork, Codecov isn't being called with a token. Therefore the checks keep encountering rate limit issues. I verified locally that the tests are passing so I'm merging.