Add advanced context aggregator with heterogeneous tree structure

[Ariq369]

Description
This PR introduces an AI powered context aggregation system for distributed trace debugging. Key features:

Main Components

Heterogeneous Tree Structure: It holds spans and logs under a single tree
ContextNode & DebugContext: Data models for structured debugging context
LLM Integration: AI driven optional feature and structure filtering
Pattern Detection: Error pattern detection automatically

Features

Multi-source aggregation: Spans, logs, and context from GitHub
LLM-driven filtering: Smartsignal's predictive maintenance platform is a multires
Performance indicators: Automatic discovery of critical indicators
Analysis of mistakes: Root cause detection with AI pattern recognition

📁 Files Added

• ai/context/models.py - Data models (ContextNode, DebugContext)
• ai/context/aggregator.py - Main aggregation logic
• ai/context/utils.py - Tree utility functions
• ai/context/__init__.py - Module exports
• ai/__init__.py - Package exports

Use

from traceroot.ai import ContextAggregator

aggregator = ContextAggregator(llm_client=my_llm_client)
debug_context = await aggregator.aggregate_context(trace_data)

[zechengz]

Thanks for the contribution.

1. Now seems a weird nested folder structure can we remove it?
2. For now let's put everything in the agent folder and rename file names properly

[Ariq369]

Thanks for the contribution.

1. Now seems a weird nested folder structure can we remove it?
2. For now let's put everything in the agent folder and rename file names properly

Yeah it's kinda messy yeah I think we should remove it and alright I'm gonna put everything in the agent folder and how do you want me to rename the files?