Research and implement non-standard RPC endpoints (debug, trace, etc.)

[LeoPatOZ]

Summary

Non-standard RPC methods such as debug_ and trace_ namespaces are not currently supported.

Problem

These endpoints are useful for debugging, tracing, and advanced node interactions but vary significantly between clients (Geth, Reth, Erigon, Nethermind, etc.).

Proposed Solution

• Research which non-standard namespaces exist across major clients
• Document client-specific differences
• Decide on a strategy: support common subset, or client-specific feature flags
• Implement chosen endpoints with appropriate documentation noting client compatibility
• Add tests (may require client-specific test environments)

Open Questions

• Which clients do we prioritize?
• How do we handle methods that exist on some clients but not others?
• Should these be behind feature flags to avoid bloating the core crate?