Skip to content
ChainFoundry

Blog

Notes from the foundry.

Engineering deep-dives, ecosystem theses, and architectural choices. Every post ties to something already shipped on crates.io — or clearly labelled as draft.

Editorial roadmap

What we're writing next.

These are drafts — not published yet. Every topic is tied to something already shipped in the chainkit repo, so the writing stays technical.

Get notified as they land

We'll send your address to info@ai2innovate.io. No third parties. Unsubscribe anytime.

  1. #04Origin8 min

    Why we built ChainFoundry

    Seven incompatible architectures, no universal toolkit. The USB-C analogy. What we've shipped on crates.io, and what's next.

    Draft · not yet published
  2. #05Deep dive12 min

    7 architectures, not 7 chains — why chain count is a misleading metric

    Ethereum vs. its 200+ EVM forks. Why SubQuery's “300 chains” and ChainFoundry's “7 families” are comparable numbers.

    Draft · not yet published
  3. #06Engineering10 min

    Building an MCP server for blockchain data

    Dual-mode (Rust PyO3 + Python fallback). Claude Desktop and Cursor integration. Tool schema design for multichain agents.

    Draft · not yet published
  4. #07Architecture9 min

    CSDL — a portable schema format for blockchain events

    Why The Graph's subgraph manifests and Envio's configs are framework-locked. How CSDL works. How to contribute schemas.

    Draft · not yet published
  5. #08Engineering7 min

    0x08c379a0 — turning revert bytes into readable errors

    How chainerrors decodes Solidity reverts, panics, custom errors, and Solana program failures — and why your AI agent needs it.

    Draft · not yet published