Ai features suite

[SuchitraSwain]

What was wrong?

Issue #

How was it fixed?

1. built lightweight AI helpers across the stack: Kademlia peer reputation, adaptive GossipSub tuning, predictive relay/DCUtR decisions, AutoNAT confidence scoring, and pubsub anomaly detection
2. added optional AI tooling for users: LLM-ready chat assistant, end-to-end ai_features_demo.py, quick-start doc, and full Sphinx guide

Summary of approach:

1. leverage heuristics (success/failure stats, moving averages, anomaly detection) so every node gains smarter behavior without heavy ML deps
2. integrate these helpers directly into the existing components (PeerRouting, GossipSub, ReachabilityChecker/DCUtR, AutoNAT, Pubsub)
3. expose configuration switches and documentation so operators can opt in, tune, and test

To-Do
[ ] Run examples/ai_features_demo.py locally to exercise all features
[ ] Try chat demo with --ai-assistant in summary and reply modes
[ ] Monitor logs when enabling adaptive tuning or anomaly detection to fine-tune thresholds

• Clean up commit history
• Add or update documentation related to these changes
• Add entry to the release notes

Cute Animal Picture

[SuchitraSwain]

@seetadev Please review