feat: add DeepEval + E2B sandbox SWE evaluation pipeline example

[Ayush7614]

What does this PR do?

This PR adds a complete end-to-end example demonstrating how to evaluate LLM-generated code using:

• OpenAI for code & unit test generation
• E2B sandbox for secure code execution
• DeepEval (GEval) for SWE-style evaluation

It shows a full pipeline:
Prompt → LLM → Sandbox Execution → Unit Tests → DeepEval Metrics

What's included?

• test_swe_pipeline.py

  • Generates Python code using LLM
  • Executes it safely inside E2B sandbox
  • Generates unit tests via LLM
  • Re-runs tests in sandbox
  • Evaluates outputs using DeepEval GEval

• tasks.json

  • List of SWE-style tasks and expected outputs

• requirements.txt

  • Minimal dependencies required to run the pipeline

• README.md

  • Setup instructions
  • API key configuration (OpenAI + E2B)
  • Run commands and explanation of workflow

Why this is useful?

This example helps users:

• Understand how to combine DeepEval with real execution
• Validate LLM code generation safely
• Add CI-style evaluation for agent / code workflows
• Use DeepEval beyond simple text evaluation

[vercel]

@Ayush7614 is attempting to deploy a commit to the Confident AI Team on Vercel.

A member of the Team first needs to authorize it.

[greptile-apps]

PR author is not in the allowed authors list.

[Ayush7614]

cc: @penguine-ip