Vandyill

About

VandyILL is a community member and independent builder in the Radix ecosystem, known for making games and small web tools, and more recently for a set of published experiments in agent-facing Radix infrastructure. Visit the community directory to discover other contributors to RADIX.wiki.

Projects

• Doubt/it! – a bluffing board game with a free-to-play browser version (doubtit.digital).
• Radix Rolodex – a directory of Radix ecosystem projects (radixrolodex.com).
• Other web projects: quack.space and cshwrd.com.

Radix agent-economy work

His site vandyill.github.io, titled The Agentic Economy, presents a set of Radix projects published under the quackstra GitHub account. They are individual experiments rather than launched products – the two payment and capital repositories have had no commits since March 2026 – but each is documented with a stated design and, in one case, a measured result worth recording.

• x402-radix – a TypeScript implementation of the x402 payment protocol for Radix, split into client, server, facilitator and core packages, and written against the scheme_exact_radix specification by Radix core developer Stefan Telea. That is the same scheme now being prepared for formal submission to the x402 standard on the community’s behalf, which makes this an early independent implementation of it rather than a competing design.

• settlement-rights – a Scrypto blueprint for pooled capital deployment: depositors wrap XRD for a receipt NFT, builders buy an authorization badge scoped to a component in order to borrow and return that capital atomically, and the badge fees accrue back to receipt holders through a per-unit accumulator.

• groth16-scrypto – a feasibility study, and the most useful of the three, because it produced a negative answer with numbers behind it (see below).

• quack.space – described on his site as a credentialing layer binding ATProtocol decentralised identifiers to Radix identities.

Verifying zkVM proofs on Radix: measured infeasible today

The groth16-scrypto study asks whether a Radix transaction can verify a BN254 Groth16 proof – the compression target that RISC Zero and Succinct SP1 zkVMs both output, and therefore the gateway to using zero-knowledge proofs as an on-ledger primitive. The verifier it builds is correct and fits comfortably in memory and binary size; the problem is execution cost. Benchmarked off-ledger against Radix’s own cost model, a single verification is reported at roughly 149–263 million cost units against a stated 100 million cost-unit ceiling per transaction, so the study concludes that pure-WASM zkVM proof verification cannot run inside a Radix transaction as the engine stands, and that native pairing operations are a prerequisite rather than an optimisation. The measurements and the engine-constraint recon behind them are published in the repository.