Enable Automatic Instrumentation for LangChain in OpenLLMetry-JS

[nirga]

Title: Enable Automatic Instrumentation for LangChain in OpenLLMetry-JS

Description:

Currently, users of openllmetry-js are required to perform manual instrumentation when tracing LangChain applications. This is documented in our JS Force Instrumentations Guide. However, this approach is less than ideal, as it adds complexity and potential for error in user implementations.

Problem:

Automatic instrumentation for LangChain is either not functioning as expected or not fully supported, leading to a lack of out-of-the-box tracing capabilities. This makes it harder for users to trace LangChain flows efficiently and accurately without manually adding spans and context in their code.

Proposed Solution:

Investigate the current limitations with auto-instrumenting LangChain in openllmetry-js.
Explore possible configurations or changes in the LangChain or OpenTelemetry integration to facilitate seamless instrumentation.
Document findings and, if necessary, create a solution roadmap that could include modifying openllmetry-js to better support LangChain auto-instrumentation.
Acceptance Criteria:

A clear understanding of the root cause behind the manual instrumentation requirement for LangChain.
A proposed path forward with actionable steps (e.g., code changes, configuration adjustments) to enable automatic instrumentation.
Updates to the documentation reflecting any improvements or necessary configurations for LangChain users.
Additional Context:

Providing robust automatic instrumentation support for LangChain would enhance usability and adoption by minimizing the setup complexity for users. This aligns with our goal of creating a smooth, reliable tracing experience in OpenLLMetry.