Managed auth provider adapters (Clerk, Supabase, Auth0)
- Dominant language
- Go
- Stars
- 0
- Forks
- 0
- Avg merge
- 3h 18m
- Merged PRs (30d)
- 26
Description
The auth foundation shaped `auth.Resolver` so a managed provider can satisfy it, and named Clerk, Supabase and Auth0 as the conforming implementations it was designed against — but none was built, and the design deferred them explicitly (`docs/superpowers/specs/2026-08-31-auth-foundation-provider-abstraction-design.md`, *Deferred*). Adopting one is more than writing a Resolver: it needs just-in-time user provisioning on first sign-in, a decision about whether the local register/login endpoints stay or are disabled, and somewhere for provider configuration to live.
Distinct from #27, which is social OAuth sign-in as another Resolver alongside the local flow; this is replacing the identity system with a hosted one. Whether all three ship or one lands first as a proof is open.
Came out of reviewing the deferred sections of the existing specs.
Contributor guide
No contributing guide indexed for this repository
Research direction
Start with docs/superpowers/specs/2026-08-31-auth-foundation-provider-abstraction-design.md, especially the Deferred section, and inspect auth.Resolver. Compare the scope with issue #27, then clarify whether one provider or all three is intended, how first-sign-in provisioning and provider configuration should work, and whether local endpoints remain; done requires those decisions and a selected implementation scope.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- go, supabase
- Domain
- authentication, backend-api-design
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Active
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100