feat(cookbook): add llamaindex-rag-tracing-eval example with traceAI Instrumentation

[pm32900]

Summary

Adds a new cookbook example: llamaindex-rag-tracing-eval.

This is a minimal, self-contained RAG pipeline that demonstrates:

• Building a VectorStoreIndex from a local document using LlamaIndex
• Tracing all LLM calls, retrieval steps, and token usage with traceAI-llamaindex
• Attaching CONTEXT_ADHERENCE, CONTEXT_RELEVANCE, and COMPLETENESS eval tags so Future AGI automatically scores each response
• Viewing full span trees and eval scores in the Future AGI dashboard

Motivation

The existing external_gdrive_search cookbook uses LlamaIndex but requires
Google Drive credentials, a database, and Gradio. There is currently no
minimal LlamaIndex RAG example that shows the tracing + eval pattern in
isolation. This fills that gap for developers who want to add observability
to a simple RAG pipeline before scaling to production.

Files Added

llamaindex-rag-tracing-eval/ ├── README.md ├── app.py ├── eval_tags.py ├── requirements.txt ├── .env.example └── data/ └── sample_policy.txt

Testing

Tested locally with Python 3.11. Run python app.py after filling in .env.
Traces and eval scores confirmed visible in the Future AGI dashboard.

Checklist

• New folder only — no existing files modified
• requirements.txt uses real published package names from PyPI
• .env.example contains no real secrets
• README includes prerequisites, install, run, expected output, and troubleshooting
• Follows the same structure as multi_agent and external_gdrive_search

[pm32900]

Hi! Just checking in to see if there's anything I can update or clarify to help move this PR forward. Happy to make any requested changes. Thanks!