Are we over-engineering multichain tooling? Need honest feedback on Chainkit? Also what are the pain-points?

[DarshanKumar89]

Hey everyone,

I’ve been building Chainkit, starting with chaincodec — a multi-chain decoding layer.

The motivation is simple:
Every time we add a new chain, we end up rewriting parsing logic.

EVM → ABI
Solana → Borsh
Cosmos → Protobuf
Aptos/Sui → BCS
Polkadot → SCALE
Bitcoin → UTXO
StarkNet → Cairo ABI

Same idea (transactions + events).
Different decoding rules.

So tools like bridges, indexers, bots, analytics systems, and dashboards end up maintaining multiple decoding stacks just to normalize data.

What I’m trying to build

A modular foundation:
chaincodec → unified decoding
chainerrors → normalized error handling
chainrpc → reliable RPC layer with failover
chainindex → simple reorg-safe indexing

The goal is not another SDK.

It’s a base layer so multichain infra doesn’t get more complex every time you support a new network.

Where I need honest feedback

Is decoding actually a painful bottleneck for you?
Or is RPC reliability / indexing / finality detection a bigger problem?
Would you adopt a canonical transaction/event format?
What would make this immediately useful in your stack?
Am I missing a more fundamental pain point in multichain infra?

Brutal feedback is welcome.
If this solves nothing, I’d rather know early.

Appreciate any thoughts from infra devs.