DNS-AID Roadmap

July 30, 2026 · View on GitHub

This page describes where the project is headed. The issue tracker is the source of truth for individual work items; this document explains priorities and direction. It is reviewed at each minor release and at governance changes.

The DNS-AID protocol is specified in the IETF (draft-mozleywilliams-dnsop-dnsaid); protocol-level changes happen there, not here. This roadmap covers the reference implementation only.

Near term (next 1–2 releases)

  • Track the IETF draft. Keep the implementation aligned with each draft revision (the draft-watch workflow flags new revisions automatically); implement wire-format and behavior changes as they land.
  • OpenSSF Best Practices Silver → Gold. Execute the OpenSSF proposal: assurance case, coverage push to 90% statement / 80% branch, reproducible-build check, fuzzing, independent security review.
  • Backend parity. Bring all DNS provider backends (Route 53, Cloudflare, NS1, Cloud DNS, Infoblox BloxOne/NIOS, Akamai EdgeDNS, DDNS) to the same feature and test level, including native private-use SVCB keys and uniform error propagation.

Medium term

  • Linux Foundation onboarding. Complete the sustainability goals in MAINTAINERS.md: a maintainer from a second organization, documented project-lead succession, and an external committer with merge rights on at least one subsystem.
  • IANA registration. Move the private-use SVCB SvcParamKeys (key65400–65405) to permanently registered keys as the draft progresses, with a migration path for published records.
  • Ecosystem interoperability. Continue ARD ai-catalog interop and keep the indexer/telemetry interfaces provider-neutral so independent directories can build on the same records.

Long term

  • 1.0 release once the IETF draft reaches a stable state (working-group adoption and wire-format stability), with semantic-versioning guarantees for the public Python API and CLI.
  • Post-quantum readiness. Mature the ML-DSA-65 HTTP Message Signatures support (pqc extra) from experimental to supported as standards settle.

Non-goals

  • Defining protocol behavior (IETF territory).
  • Canonicalizing a single directory/indexer service; DNS-AID stays a substrate that any directory can build on.