Repository navigation
chore(repo): sync develop into next - #7232
Conversation
## Summary **TL;DR:** `gen types` bundles `@supabase/typegen@0.3.1`, which pins `@supabase/postgrest-typegen@0.6.0`. `--lang swift` now runs supabase-swift's `supabase-typegen` binary instead of an in-process generator, view columns that select a `NOT NULL` source column are typed non-nullable, and the introspected document gains the resolved `rpcFunctions` collection. ## What changed - Bump `@supabase/typegen` from `0.2.2` to `0.3.1` in `apps/cli/package.json`. The lockfile moves `@supabase/postgrest-typegen` from `0.4.0` to `0.6.0` and `arktype` from `2.2.5` to `2.2.8` (with `@ark/schema`, `@ark/util` and `arkregex`). - `pnpm-workspace.yaml`: the `minimumReleaseAgeExclude` entries for the two typegen packages move to the new versions, both published on 2026-10-16. The arktype entries are removed, since 2.2.8 was published on 2026-10-07 and is already past the 7-day window. - `SIDE_EFFECTS.md` for `gen types`: Swift is listed as an external tool, and `--swift-access-control` accepts only `internal` and `public`. ### Output changes - **Swift** ([supabase/sdk#264](supabase/sdk#264)): output comes from supabase-swift's `supabase-typegen`, emitting `@Table` structs for the PostgREST v3 typed API. The binary is resolved from `SUPABASE_TYPEGEN`, `PATH`, or the supabase-swift release the project depends on. `--swift-access-control private` and `package` now fail validation before anything runs. - **TypeScript, Go, Python** ([supabase/sdk#284](supabase/sdk#284)): a view column that selects a `NOT NULL` source column is typed non-nullable instead of nullable. - **TypeScript** ([supabase/sdk#272](supabase/sdk#272), [supabase/sdk#277](supabase/sdk#277)): a function whose only parameter is an unnamed row is no longer typed as callable through `/rpc`, and the computed-field fallback is narrowed. - **Dart**: the document piped to `dart run supabase_typegen --output -` now carries `rpcFunctions` ([supabase/sdk#271](supabase/sdk#271)), which supabase_typegen reads to generate typed RPC methods. ## Linked issue No GitHub issue. - [x] The linked issue is **open** and carries the `open-for-contribution` label (or I'm a Supabase maintainer). ## Checklist - [x] The PR title follows [Conventional Commits](https://www.conventionalcommits.org/) (e.g. `fix(cli): …`). - [x] Tests added or updated for the change. No test in this repository asserts on generated output; the generators' own tests cover it in supabase/sdk. - [x] `types:check`, `lint:check`, `fmt:check`, `knip:check` and `lint:effect:check` pass across the workspace. `lint:actions:check` and the Go lint were not run locally, and this PR touches no workflow or Go code. `gen` tests on bun 1.4.2: 114 unit tests pass (1 skipped), 196 integration tests pass.
This PR was automatically created to sync the generated `@supabase/api` package with the latest Management API OpenAPI document. Changes were detected in the upstream OpenAPI documents exposed by `https://api.supabase.com/api/v1-json` and `https://api.supabase.com/api/v2-json`. Co-authored-by: jgoux <1443499+jgoux@users.noreply.github.com>
# Conflicts: # apps/cli/src/commands/gen/types/SIDE_EFFECTS.md
Claude resolved the merge of
|
| File | Resolution |
|---|---|
apps/cli/src/commands/gen/types/SIDE_EFFECTS.md |
Kept next's versions of the Files Read, Subprocesses and Environment Variables tables: config.json/config.toml [db] reading, the docker daemon note, and the new SUPABASE_API_SCHEMAS/SUPABASE_DB_PASSWORD/SUPABASE_SERVICES_HOSTNAME/SUPABASE_WORKDIR wording. Next's removal of the SUPABASE_EXPERIMENTAL_ORIOLEDB_VERSION row is kept. Into those tables I added develop's Package.resolved row, the supabase-typegen --access-control subprocess row, and the SUPABASE_TYPEGEN and XDG_CACHE_HOME/HOME env rows; develop left their Required? cell empty, so I filled it with no. In the output-compatibility note I took develop's line, which says Swift output now comes from supabase-swift's supabase-typegen, over next's rewording of the old 'byte-identical' line. |
Decisions for @supabase/cli. Confirm them, or push a change to this branch:
- The Go/Swift output-compatibility note: develop says Swift now comes from supabase-swift's generator, while next only reworded the old claim that Swift output is byte-identical. Which should stay? Chose: Develop's text, because it describes the current Swift generator (supabase-swift's
supabase-typegen); next's wording would bring back a claim that is no longer true.
Quick checks. The merged tree passes check:all and unit tests.
Approve to fast-forward next once checks pass (workflow run).
There was a problem hiding this comment.
🤖 AI Review
Both independent reviews completed and reported no findings. Verification of the checked-out documentation resolution and related code found no sync-introduced defects or cross-branch incompatibilities.
Findings
No issues found.
Stats
Claude findings: 0 · Codex findings: 0 · Confirmed: 0 · Refuted: 0 · Uncertain: 0
Models: claude-opus-5-5 + gpt-6.1-sol · Trigger: manual · Workflow run
This review runs once per PR. A maintainer can request another with a /ai-review comment.
supabase-oss
left a comment
There was a problem hiding this comment.
CI is green and the AI review raised nothing the merge introduced, so this approves Claude's resolution of the sync (https://github.com/supabase/cli/actions/runs/39233364236).
Merging
developintonextconflicts. This branch starts at the tip ofdevelop(dca88f9);nextis at8851a28.Conflicting files:
apps/cli/src/commands/gen/types/SIDE_EFFECTS.mdTo resolve:
git fetch origin git switch sync/develop-into-next git merge origin/next # resolve the conflicts, then commit git push origin sync/develop-into-nextApprove to fast-forward
next; do not use the merge button.