We build trustless resolution architectures and reference implementations for cross-namespace identity systems.

• AI can generate convincing artifacts at infinite scale
• Identity, authorship, and provenance are easily forged
• URLs, usernames, and "official" sources are no longer trustworthy by default

While generation is abundant, verification is scarce — our work exactly addresses this asymmetry.

We've identified Managed Agent Runtime Platforms (MARPs) as the next operator class. By enabling a new payload type — authority-policy records — and the verification path for it, we let agents transact reliably across MARPs.

ENS is the natural substrate for this emerging operator class, but it currently lacks a standardized, resolver-native "agent trust runtime" for dynamic control, validity, permissions, and portable reputation.

Solving identity-shaped problems is how ENS has proven useful and grown before — the canonical case being Coinbase issuing cb.id subnames at scale.

By naming MARPs as the new operator class, we're building the verification-and-revocation toolkit that lets developers turn subname issuance into Managed Agent Identity Platforms (MAIPs) for those MARPs — making ENS the open, vendor-neutral authority substrate for the agent economy.

The design principle is simple: data at the edge, trust at the root.

We're productionizing the Authority tier of the MAIP stack: a Verifier + per-name AuthResolver, a TypeScript SDK, a conformance suite, and integration guides that let any service confirm — in real time — whether an action attributed to an ENS-named agent is currently authorized, and whether the credential behind it has been rotated, expired, or revoked.

Built on ENSv2 primitives; complements core ENS Labs work.

• Thesis: The next operator class — Managed Agent Runtime Platforms
• Application + spec: [steg-eth/spp3](https://github.com/steg-eth/spp3)
• Read-only pilot (record-publishing + resolution path), draft PR: [BankrBot/skills#189](https://github.com/BankrBot/skills/pull/189) · verified run log

A separate, in-flight DAO proposal — not part of SPP3. Programmable TLD assignment so post-2012 gTLDs can claim ENS names under DAO-defined policy, shipped as a two-phase rollout: Phase 1 deploys with execution disabled; the DAO flips executionEnabled in Phase 2 after ratifying review procedures.

• RFC · Temp Check: Authorize TLDMinter as Root Controller
• [dao-proposals](https://github.com/steg-eth/dao-proposals) — executable proposals: calldata, tests, rationale
• [dao-audits](https://github.com/steg-eth/dao-audits) — independent bytecode-verification artifacts
• [dnssec-solutions](https://github.com/steg-eth/dnssec-solutions) — EIP-7951 precompile demonstrations

We contribute to canonical ENS contracts: [ensdomains/ens-contracts#509](https://github.com/ensdomains/ens-contracts/pull/509), merged and shipped in v1.7.0. Our fork: [steg-eth/ens-contracts](https://github.com/steg-eth/ens-contracts).

• **estmcmxci.eth** — ENS architecture, specifications, and ecosystem integration.
• **mouzayan.eth** — protocol engineering, smart-contract delivery, and security.

• Delegates — reach out for a walkthrough or co-sponsorship on any proposal.
• Builders — follow the specs and demos, open issues, or co-own integrations.
• Tag **@estmcmxci.eth** on the ENS forum.