mirror of
https://github.com/supabase/supabase.git
synced 2026-10-08 19:05:06 +03:00
docs/ssr-reference
4862
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 Danny WhiteandJoshen Lim
|
b82dec4ef1 |
fix(studio): polish advisor attention indicator quirks (#47714)
## I have read the [CONTRIBUTING.md](https://github.com/supabase/supabase/blob/master/CONTRIBUTING.md) file. YES ## What kind of change does this PR introduce? Bug fix / UI polish. ## What is the current behavior? - The Advisors sidebar attention dot drifts 1–2px sideways on hover-expand, because its `left` offset was tied to the expanded state while the nav icon only shifts when the sidebar is persistently open. - When the Advisor Center header button is selected while in a critical state, it keeps a destructive outline instead of matching the other header circles (`bg-foreground`). ## What is the new behavior? - Sidebar attention dot offset follows the same condition as nav icon padding (persistently open), so it no longer drifts on hover-expand. - Selected Advisor Center button matches the other header circles (foreground fill, no destructive outline). Critical idle styling is unchanged aside from a destructive hover border. The critical dot is slightly lighter when selected in light mode so it still contrasts on the inverted fill. ## Additional context Earlier commits on this branch experimented with a shared `useAdvisorAttention` hook to sync the sidebar and header indicators. That was dropped: the sidebar Advisors route goes to project security/performance pages, while the header opens Advisor Center (including org notifications). Those surfaces should not share one attention definition — thanks Joshen for catching that. --------- Co-authored-by: Joshen Lim <joshenlimek@gmail.com> |
||
|
Ali Waseem
|
620c9329cf |
fix(studio): prevent duplicate success toast on disk upgrade (#48097)
## Summary Fixes FE-3948: two success toasts were firing after a disk upgrade. Now the immediate toast is skipped when a disk resize is requested, since the polling effect already shows a completion toast once the resize is applied. <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Bug Fixes** * Improved disk settings update notifications to prevent premature success messages while configuration changes are still being applied. * Success confirmation now appears after disk resizing is fully completed. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Etienne Stalmans
|
05d5da0340 |
chore: add ssl_enforcement_required for jit (#48032)
## I have read the [CONTRIBUTING.md](https://github.com/supabase/supabase/blob/master/CONTRIBUTING.md) file. YES ## What kind of change does this PR introduce? chore / bug fix ## What is the current behavior? The new `ssl_enforcement_required` state is not handled ## What is the new behavior? Displays the correct message <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Bug Fixes** * Improved temporary database access messaging when SSL enforcement is required. * Added a direct action to open database settings and enable SSL enforcement before activating temporary access. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Joshen Lim
|
13659ecc50 |
Chore/refactor observability layout menu (#48089)
## Context Just refactors `ObservabilityMenu` to retrieve the menu items via a hook + scaffold the Top for Postgres menu item No functional changes here <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added a keyboard shortcut (`U`, then `C`) for quickly opening Observability Connections. * Improved observability navigation with feature-aware sections and consistent URL parameter preservation. * Custom reports are now sorted alphabetically and include available actions directly in the menu. * **Bug Fixes** * Improved handling of missing report details and duplicate or unsupported query parameters. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Jordan McQueenandJoshen Lim
|
fa1e4c4bbf |
feat(studio): add ClickHouse replication destination (#46870)
Adds ClickHouse as a replication destination type in Studio.
- New ClickHouse option in the destination type selector, gated behind
the
`etlEnableClickHousePrivateAlpha` organization feature flag (off by
default).
- ClickHouse settings form: URL, user, password (optional), database,
and
- Client-side URL validation requires HTTPS and rejects URLs targeting
internal addresses (loopback, RFC 1918, link-local, CGNAT, IPv6
loopback/link-local/ULA, and IPv4-mapped/NAT64 forms). Server-side
validation remains authoritative.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
## New Features
- Added **ClickHouse** as a replication destination option (private
alpha), including support in destination selection/panel, replication
diagram rendering, and destination icons.
- Introduced a ClickHouse destination form with fields for URL, user,
optional password (masked toggle), database, and engine selection.
- Added ClickHouse destination config handling for create/update flows,
with normalization and engine support.
## Tests
- Expanded unit tests to cover ClickHouse validation and destination
config building/normalization, including HTTPS-only and blocking
localhost/internal targets.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: Joshen Lim <joshenlimek@gmail.com>
|
||
|
Joshen Lim
|
8962308215 |
Add support for multiple custom auth providers in custom-content (#48030)
## Context Adds support for multiple custom auth providers in custom-content <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added support for multiple custom sign-in providers via a new plural configuration. * Updated the sign-in page to render all configured custom provider options while maintaining compatibility with the legacy single-provider setting. * Improved the custom provider button display to remove internal prefixes from provider names. * **Documentation** * Updated the configuration schema, examples, and sample data to document the new multi-provider setting. * Marked the legacy single-provider configuration as deprecated in favor of the plural option. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Jordi Enric
|
fc72a6b259 | fix(studio): preserve API report filter value casing so method filters match (#48056) | ||
|
Danny White
|
50e9fedb20 |
feat(studio): finesse logs date picker range colours (#48019)
## What kind of change does this PR introduce? UI polish / bug fix for the shared Calendar range selection and Logs date picker. ## What is the current behavior? - Selected date ranges use opaque `brand-400` / `brand-500` fills that read too loud in light mode, with black text that is hard to read on darker endpoints. - Start/end days are squared off on the connecting edge without intentional outer rounding. - Outside days in a selected range are dimmed with `opacity-50`, which can tint the range wash incorrectly when a range starts in the prior month. - The large-range warning in `LogsDatePicker` is a full-bleed yellow banner that feels too heavy for the popover. - Time inputs show a clock icon that adds visual noise. ## What is the new behavior? - Range middle uses a softer `brand-200` wash; start/end stay on stronger brand fills with readable foreground text. - Start days round on the left (`rounded-l-md`), end days on the right (`rounded-r-md`); day hover keeps `rounded-md`. - Selected outside days and “today” no longer fight the range wash colours. - Large-range warning is quiet inline `text-warning` copy that wraps to the calendar column width. - Clock icon removed from `TimeSplitInput`. | Before | After | | --- | --- | | <img width="1096" height="1076" alt="CleanShot 2026-07-16 at 17 59 21@2x" src="https://github.com/user-attachments/assets/eac38022-ed92-4dbe-9932-55f7bf0af934" /> | <img width="988" height="1064" alt="CleanShot 2026-07-16 at 17 59 34@2x" src="https://github.com/user-attachments/assets/3dce30dd-9fd8-4da4-82b3-5663250a4ddc" /> | ## Additional context Shared `Calendar` changes apply anywhere range mode is used, not only logs. <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit - **Style** - Refined the date-picker popover layout for start/end controls with better fit and max-width handling. - Updated calendar day and range visuals (selection, outside states, rounding, and hover behavior) to reduce “ghost” styling and improve consistency. - Restyled the large-range warning to improve spacing and alignment. - Simplified the time-splitting input UI by removing the leading clock icon. - Adjusted the “Copy range” button feedback color for copied/pasted states. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
seungjaeandGildas Garcia
|
77953e7a1f |
feat(studio): include enums and RLS policies in Schema Visualizer Cop… (#46189)
Extend the "Copy as Markdown" feature in the Schema Visualizer to include Custom Types/Enums and Row Level Security (RLS) Policies in the generated output. Closes https://github.com/orgs/supabase/discussions/46108 ## I have read the [CONTRIBUTING.md](https://github.com/supabase/supabase/blob/master/CONTRIBUTING.md) file. YES ## What kind of change does this PR introduce? Feature — extend Copy as Markdown to include enums and RLS policies ## What is the current behavior? The "Copy as Markdown" function in the Schema Visualizer generates a SCHEMA.md that only includes tables, columns, and entity relationships. Custom Types/Enums and RLS Policies are missing, making the output an incomplete representation of the schema. Related: https://github.com/orgs/supabase/discussions/46108 ## What is the new behavior? The generated markdown now includes two additional sections: **Custom Types / Enums:** - Lists each enum type with its ordered permitted values - Filtered by the currently selected schema **RLS Policies:** - Grouped by table for self-contained readability - Includes policy name, command (SELECT/INSERT/UPDATE/DELETE/ALL), roles, action (PERMISSIVE/RESTRICTIVE), USING expression, and WITH CHECK expression **Example output:** ```markdown ## Custom Types / Enums ### `order_status` `pending` | `processing` | `shipped` | `delivered` ## RLS Policies ### `orders` | Policy | Command | Roles | Action | USING | WITH CHECK | |--------|---------|-------|--------|-------|------------| | `users_own_orders` | SELECT | authenticated | PERMISSIVE | `auth.uid() = user_id` | — | ``` ## Additional context - 3 files changed: `Schemas.utils.ts`, `SchemaGraph.tsx`, `Schemas.utils.test.ts` - New utility functions `getEnumsAsMarkdown()` and `getPoliciesAsMarkdown()` with unit tests - Reuses existing `useEnumeratedTypesQuery` and `useDatabasePoliciesQuery` hooks - No breaking changes — existing markdown output is preserved, new sections are appended <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * "Copy as Markdown" now includes enumerated types and database policies alongside existing schema/table content; policies are included with their rule details and grouped by table. * **Tests** * Added tests covering enum and policy markdown generation, including matching/non-matching schema cases and policy formatting. <!-- review_stack_entry_start --> [](https://app.coderabbit.ai/change-stack/supabase/supabase/pull/46189?utm_source=github_walkthrough&utm_medium=github&utm_campaign=change_stack) <!-- review_stack_entry_end --> <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Co-authored-by: Gildas Garcia <1122076+djhi@users.noreply.github.com> |
||
|
Gildas Garcia
|
dba31df91d |
fix: scoped PAT creation form error messages are hidden (#48011)
## Problem When creating a new scoped PAT, if users didn't add at least one permission or have a misconfigured permission (no access selected), the form does not submit but no error message is shown. The UI looks broken. ## Solution This is because there's a zod validation happening but its messages are not displayed for permissions. The proper fix is to use react-hook-form field array. <img width="541" height="633" alt="image" src="https://github.com/user-attachments/assets/89cab58d-761e-4131-9bce-460625067f8a" /> <img width="540" height="594" alt="image" src="https://github.com/user-attachments/assets/ed95cee0-06b5-4233-9a23-6819fb0e1a17" /> <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Bug Fixes** * Improved permission selection and toggling behavior in the scoped access token flow. * Enhanced validation feedback for permission rows and action selections, keeping error states in sync after changes. * Updated error handling to surface permission-related messages more reliably. * **Refactor** * Reworked the permissions UI to use a more reliable control-based rendering approach for rows, selection changes, and error presentation. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Ivan Vasilov
|
dc3c8684cc |
chore(deps): upgrade valtio to v2 (#48031)
Audited all proxy()/useSnapshot() usage against the v1→v2 migration guide; no breaking changes apply (no reused proxy() inputs, no promise-valued state, all consumers already client components). <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Chores** * Updated the Valtio dependency to a newer version for improved compatibility. * **Bug Fixes** * Improved AI assistant persistence in IndexedDB so chat sessions reliably save (while keeping only the most recent 20 messages per chat). * Hardened tabs restoration from storage to fall back to fresh defaults when data is missing, invalid, or fails validation. * **Refactor** * Switched multiple studio panels to use fresh initial-state factories for initialization and reset reliability. * Updated advisor state so the derived notification filter count is no longer exposed. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Riccardo Busetti
|
f19cb09b43 | feat(pipelines): Update product docs and UI copy (#47997) | ||
|
|
3c6ef31959 |
feat: add User Filter to the unified logs (#47879)
Offshoot from https://github.com/supabase/supabase/pull/47743. [Linear issue](https://linear.app/supabase/issue/FE-3939/add-user-logs-filter-to-the-logs-page) <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added a `user` filter to Unified Logs with a user picker (email or ID). * Added “View user logs” actions from the Users table to jump to Unified Logs. * **Bug Fixes** * Updated Unified Logs searching so default log-type restrictions no longer block user-attributed results. * **UI Updates** * Unified Logs filter bar and reset behavior now include clearing the user filter. * Improved empty-state messaging when the selected user filter isn’t supported. * Refreshed highlighted styling in command list items. * **Tests** * Expanded coverage for user filter configuration and query edge cases. <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Co-authored-by: kemal.earth <606977+kemaldotearth@users.noreply.github.com> Co-authored-by: kemal <hello@kemal.earth> Co-authored-by: Joshen Lim <joshenlimek@gmail.com> |
||
|
Jordi Enric
|
18b273cde0 |
fix(studio): hide Postgres Version Upgrade logs behind a flag (#48007)
## Problem The legacy Logs Explorer's Postgres Version Upgrade page (Database Operations section) shows no results for most users. pg_upgrade_logs rows aren't tagged with the project attribute the shared logs endpoint scopes on, only a host field, so the project-scoped query returns nothing even though the data exists. ## Fix Hides the Database Operations sidebar section behind a new showPostgresUpgradeLogs feature flag (default off), following the same pattern as showMultigresLogs. The page route itself is untouched, only the sidebar entry point is gated. ## How to test - With the flag off, open the legacy Logs Explorer sidebar and confirm the Database Operations section and Postgres Version Upgrade item are gone - With the flag on, confirm the section and item render as before <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added conditional access to Postgres Version Upgrade logs based on feature availability. * Failed Postgres upgrade notifications now show a “View logs” option only when supported. * **Bug Fixes** * Prevented unsupported Postgres upgrade log links and navigation options from appearing. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Joshen Lim
|
6e0078182b |
Consolidate project creation UI for vercel integration flow (#47995)
## Context There's 2 areas of the dashboard that has the project creation flow - and this PR consolidates both to use the same UI components to minimise duplication + keep things consistent ### Before <img width="1920" height="957" alt="image" src="https://github.com/user-attachments/assets/2a7ab79d-71c7-43f2-925b-1e1666cc3a69" /> ### After <img width="1389" height="957" alt="image" src="https://github.com/user-attachments/assets/f8465568-af99-46eb-80ee-7ac345383231" /> ## Changes involved - What this means for the project creation flow for Vercel Integration: - Smart region can be selected - Compute size can be selected - Enable Data API can be checked - Automatic RLS enable can be checked - How it differs from the main project creation flow on `new/slug` - Organization selection is disabled (cannot be changed) - The following UI is hidden: - "Internal configuration" section - "GitHub repository" field - "Free project info" at the bottom - "Cancel" button Eventually we could looking into reducing the differences more, e.g having data seeding for both ways, and showing GitHub repository field for Vercel integration Resolves DEPR-616 Resolves FE-3905 ## To test Tbh, I'm not really sure how you'd be able to test the vercel integration locally or on staging, this seemingly can only be done when changes land on prod. - What I'd do however is to just test the project creation flow minimally by landing on `/integrations/vercel/_/deploy-button/new-project` - Project creation can work, but just not the connection creation part - And also test project creation on `/new/slug` as well <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit ## Summary by CodeRabbit * **New Features** * Added “Create sample tables with seed data” during project creation. * Enhanced Vercel integration setup with a guided creation flow and post-creation connection step. * Added an option to disable organization selection in specialized flows. * Added support for triggering a callback after successful project creation. * Added support for hiding the Cancel button in specialized flows. * **UI Improvements** * Refined the connected GitHub repository selector button/dropdown visuals. * Improved security options behavior for different project creation contexts. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Charis
|
1c827c5cbb |
refactor(sql-editor): extract title-gen/execute-params + merge auto-limit functions (#48013)
## Summary Part 2/6 of the SQL Editor testability follow-up, stacked on #47980 (the analyzeQueryIssues/resolveConnectionString PR). - Extracts `shouldAutoGenerateTitle` and `buildExecuteParams` out of `useSqlEditorExecution`'s inline logic into `SQLEditor.utils.ts`. - Merges `checkIfAppendLimitRequired` and `suffixWithLimit` into a single `applyAutoLimit` function — the two were only ever called together and re-parsed the same query twice at every call site. `applyAutoLimit` only accepts `SafeSqlFragment` (never a plain string) and composes the `LIMIT` suffix through `safeSql`/`literal` rather than raw template concatenation, so the only place in the file that reasserts the `SafeSqlFragment` brand on a derived string is the small, dedicated `trimTrailingSemicolons` helper — removing existing terminators can't introduce unsafe content, unlike gluing new text onto the fragment. - Updates the two other `checkIfAppendLimitRequired`/`suffixWithLimit` call sites (`EditorPanel.tsx`, `ReportBlock.tsx`) accordingly; `ReportBlock` now promotes its report SQL once and reuses the result for both its display-only auto-limit hint and its execution, instead of promoting twice. ## Test plan - [x] `pnpm --filter studio typecheck` - [x] `pnpm test:studio -- SQLEditor ReportBlock EditorPanel` (239 tests passing) |
||
|
Charis
|
11fb715149 |
refactor(sql-editor): extract query-issue analysis + connection string resolution (#47980)
## Summary Part 1/6 of the SQL Editor testability follow-up (extracting pure decision logic out of the SQL editor hooks so it can be exhaustively unit-tested without mocking). - Extracts `analyzeQueryIssues` and `hasBlockingIssues` out of `useSqlEditorExecution`'s inline destructive-query / warning-modal-gating logic into `SQLEditor.utils.ts`. - Extracts `resolveConnectionString`, deduping the `databases?.find(...)` lookup that was duplicated verbatim in both `useSqlEditorExecution` and `useSqlEditorExplain`. - Behavior-preserving — same runtime logic, now unit-testable in isolation. ## Test plan - [x] `pnpm --filter studio typecheck` - [x] `pnpm test:studio -- SQLEditor.utils` (159 tests passing, includes new exhaustive-permutation cases for the three extracted functions) <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Bug Fixes** * Improved SQL safety checks before execution, including destructive queries, unsafe updates, database-altering commands, and tables missing row-level security. * Correctly recognizes tables protected by active security triggers. * Improved database connection selection for primary and read-replica databases. * Preserved the ability to run queries with force enabled when safety warnings are present. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Jordi Enric
|
9069e8d21e |
fix(reports): chart click selects bucket range, fix logs datepicker crash (#48009)
## Problem In a database report, dragging over a chart to open logs had two bugs. Clicking a bar instead of dragging produced a zero-width selection (its === ite), so the logs view showed no results. Separately, the logs datepicker crashed with "Invalid time value" because the chart put raw epoch-ms values into the its/ite URL params, which get parsed with new Date(string) and yield an Invalid Date that react-day-picker cannot format. ## Fix The chart now emits ISO timestamps in the logs URL so its/ite match the format every other logs consumer already uses, and the datepicker guards against unparseable values from any source (such as old bookmarked links). A single click now expands the highlight to the clicked bucket, from the bar's start to the next bar's start, so the selection, popover, and logs range all cover the bar the user picked. ## How to test 1. Open a database report with a chart (for example Postgres, project logs). 2. Drag across the chart, choose "Open in Postgres Logs", then open the timepicker. Expected: it opens without crashing and shows the selected range. 3. Go back and single-click one bar instead of dragging. Expected: that bar's bucket is selected and "Open in Postgres Logs" loads logs for a real, non-empty range. 4. Run the chart tests, expected all pass: pnpm --filter studio exec vitest --run components/ui/Charts <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Bug Fixes** * Prevented invalid log date inputs from creating invalid date selections. * Improved chart “open logs” links by normalizing the selected time bounds (now consistently sent as ISO timestamps). * Refined chart highlight behavior for click-to-advance and more accurate left/right range selection, supporting both numeric and string coordinate values. * Updated the chart highlight dropdown display for clearer formatting of numeric vs non-numeric dates. * **Tests** * Added Vitest coverage for chart highlight click, drag, and left/right ordering scenarios. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
ChloeGarciaMillerand
|
f29db88edd |
add aria label in Query Performance (#47678)
## I have read the [CONTRIBUTING.md](https://github.com/supabase/supabase/blob/master/CONTRIBUTING.md) file. YES ## What kind of change does this PR introduce? Changes to the structure of the <Switch> and <Dialog> elements, which previously resulted in two nested buttons in the DOM. Adding the asChild attribute broke the styling of the <Switch>. Added aria-label and tooltips to the <Switch> component and the icon-only buttons. <img width="1909" height="1254" alt="fix_observabilty" src="https://github.com/user-attachments/assets/4dc99845-c083-455b-beed-1ac58b66dd6f" /> ## What is the current behavior? Icon-only buttond do not have explicit accessible names for screen readers. Two nested buttons are invalid in HTML and interfere with the operation of screen readers. ## What is the new behavior? Iicon-only buttons now have explicit accessible names using visually hidden text (sr-only), ensuring proper screen reader support. The DOM structure is now correct. ## Additional context No visual changes were introduced. <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit ## Summary by CodeRabbit * **Bug Fixes** * Improved accessibility in the query performance filter bar by adding clear `aria-label` values to the Refresh and Reset report buttons. * Improved accessibility and usability for cron job table actions by adding a job-specific “actions” tooltip/label for the dropdown trigger. * Updated the cron job enable/disable toggle with job-specific `aria-label`s and refined the confirmation dialog flow to prevent unintended row interactions. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Vaibhav
|
05d8dd356c | fix: selfhost edge URLs (#47861) | ||
|
David Whittington
|
591b621567 |
fix(studio): tighten time bounds on single-log inspection query (#47978)
## Summary - The unified log inspection point-lookup (`getUnifiedLogInspection` in `apps/studio/data/logs/unified-log-inspection-query.ts`, used by `ServiceFlowPanel` when a user selects a row to view its detail panel) previously reused the whole selected search date range for its `iso_timestamp_start`/`iso_timestamp_end` bounds, even though it looks up exactly one row by `id`. With a wide search range selected (days/weeks), this scans far more of the ClickHouse-backed `logs` table than necessary. - Since the selected row's own timestamp is already known client-side, the query now bounds itself to a ±1 hour window around that timestamp instead, falling back to the previous search-range behavior when no timestamp is available. - No SQL text changes for the time bound — the `iso_timestamp_start`/`iso_timestamp_end` params are the existing mechanism by which every other query in this file (and sibling logs queries) bounds time server-side, so this follows that same convention rather than adding a redundant inline `WHERE timestamp` clause. - Also added an explicit `AND source = '...'` filter to the OTEL point-lookup SQL. The logs table's primary key is `(project, source, timestamp)`, so filtering on `source` narrows the sorted range before the timestamp bound even applies — the service flow `type` already maps 1:1 to a `source` value, so no new data was needed at the call site. ## Test plan - [ ] Typecheck (couldn't run locally in this environment — no `node_modules` installed) - [ ] Manually verify in Studio: open Logs Explorer with a wide time range (e.g. 7 days), select a log row, confirm the detail/service-flow panel still loads the correct enriched data - [ ] Confirm behavior is unchanged when `logTimestampMs` is unavailable (falls back to search range) 🤖 Generated with [Claude Code](https://claude.com/claude-code) <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Bug Fixes** * Improved unified log inspection accuracy by narrowing the inspection window to ±1 minute around the selected log event when a timestamp is available. * Updated service-flow and OTEL inspection lookups to use the selected log entry’s timestamp for tighter, more relevant results. * Preserved the prior broader time-range behavior when a timestamp isn’t available. * **Refactor** * Centralized log type → source mapping and generated the corresponding query filters from that shared mapping for consistency. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Joshen Lim
|
22b3419a28 |
Extract project creation form into its own component (#47957)
## Context This is just a pre-requisite to consolidating the project creation UI as there's another page that has the project creation flow too [here](https://github.com/supabase/supabase/blob/master/apps/studio/pages/integrations/vercel/%5Bslug%5D/deploy-button/new-project.tsx). So the next step will just be to use the same `ProjectCreationForm` there No functional changes here - just moving things around ## To test - [ ] Verify that project creation still works <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added a full “create project” experience with eligibility-aware defaults, advanced configuration sections, optional GitHub integration, and compute-cost confirmation when applicable. * **Improvements** * Enhanced project-creation success/error handling and navigation. * Refined CLI backup/restore dialogs (better layout/wording, accessibility updates, and improved section separation). * **Documentation** * Standardized all relevant documentation links across the app using a shared `DOCS_URL` source. * **Refactor** * Refactored the “New Project” page to delegate the wizard UI and flow to a reusable creation component. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Ali WaseemandJoshen Lim
|
ae957414b4 |
fix(studio): make unified logs sidebar banner dismissible (#47977)
## Summary - Adds a close button to the "Introducing unified logs" sidebar banner, storing the dismissal in localStorage so it stays hidden. ## Test plan - [ ] Open Logs Explorer, confirm the X dismisses the banner and it stays gone after reload. <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added an `X` close button to the unified logs preview banner. * Remember banner dismissal using local storage, so it stays hidden after closing. * Updated banner visibility rules to account for unified-logs preview enablement and default opt-in state. <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Co-authored-by: Joshen Lim <joshenlimek@gmail.com> |
||
|
Ali Waseem
|
b100272376 |
chore(sql-editor): remove Pretty Explain feature (#47981)
Removes the SQL Editor Pretty Explain feature — the Explain tab, the Run EXPLAIN ANALYZE action + shortcut, and its dead plumbing. It's been gated off behind the `DisablePrettyExplainOnSqlEditor` kill switch for weeks with no usage or complaints. `ExplainVisualizer` / `isExplainQuery` are kept — they're used independently by Query Insights, Query Performance, and the EditorPanel quick-runner. Manually-run `EXPLAIN` queries still render as raw rows in the Results tab. Typecheck, lint, and all affected unit tests pass. Closes FE-3930 <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Changes** * Removed the SQL editor’s EXPLAIN execution workflow, including its toolbar action, keyboard shortcut, utility tab, and visual query-plan display. * Simplified query execution to focus on standard results and charts. * Improved result clearing when switching databases and refined execution error handling. * Updated SQL editor state and tests to reflect the streamlined experience. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Charis
|
4df3c86e93 |
feat: add admin studio link (#47976)
Add link to Admin Studio in top bar for internal use only. Resolves FE-3864 <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added a conditional Admin Studio shortcut to the desktop project header. * The shortcut opens the relevant project in Admin Studio and is available only when configured and enabled. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Charis
|
9e80a159b5 |
refactor(sql-editor): decompose into controller contexts + presentational components (decompose 6/6) (#47938)
## What Final step (**6 of 6**) of the `SQLEditor.tsx` decomposition. Splits the two `ResizablePanel` bodies out of the `SQLEditorContent` monolith into presentational sub-components: - **`SQLEditorPane`** — the editor panel: loading state, `DiffEditor` + diff ask-AI widget, `MonacoEditor` + ask-AI widget. Reads the Monaco refs (`editorRef`/`monacoRef`/`diffEditorRef`) from `SQLEditorContext`, and receives the `diff`/`prompt`/`ai` controllers plus reactive state as props. - **`SQLEditorResults`** — the results panel: loading state + `UtilityPanel`. `SQLEditorContent` is now a composition root: shared-ref provider, hook composition, and the run-query warning modal. ## Result `SQLEditor.tsx` goes from the original **1056-line** monolith down to **259 lines** (≈75% reduction). The remaining size over a bare ~140-line root is the run-query warning modal, kept inline **deliberately**: its handlers hold the `acceptUntrustedSql` promotion, which must stay at the explicit user-action boundary in the root rather than moving into a presentational pane. ## Behavior-preserving The moved JSX is byte-identical aside from prop threading. No logic, effects, dependency arrays, or `eslint-disable`s changed. The two **render-time ref reads** — the editor placeholder (`!promptState.isOpen && !editorRef.current?.getValue()`) and the ask-AI widget gate (`editorRef.current && promptState.isOpen && !isDiffOpen`) — are preserved verbatim in `SQLEditorPane`, which re-renders whenever the `prompt`/`diff` props change, keeping those reads fresh (the guardrail from the plan). Verification (all green): - `SQLEditor.test.tsx` characterization suite — 11/11 pass - `tsc --noEmit` — no new errors - eslint — clean (no new ratchet entries) - prettier — clean ## Stack Builds on decompose 5 (#47935). This is the last PR in the series — the decomposition is complete after this merges. <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added a resizable SQL editor layout with run warnings, query results, and an Explain view. * Introduced centralized SQL editor controllers and expanded AI-assisted prompt/diff workflows. * Added keyboard support for running and Explain analysis. * **Bug Fixes** * Restored editor focus after accepting or discarding AI changes. * Improved handling and validation of SQL Explain actions. * Preserved editor scroll position when switching snippets. * **Refactor** * Streamlined the SQLEditor into a composed layout and improved memoization to reduce unnecessary re-renders. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Danny White
|
2279c4e69e |
fix(studio,www): repair Join us on Discord buttons (#47711)
## What kind of change does this PR introduce? Bug fix ## What is the current behavior? The "Join us on Discord" buttons are broken in two places: - **Studio support sidebar** (`HelpPanel`): on light mode the button background is not white, so the intended white button with purple icon and text is broken. - **www `/support` page** (shared `secondary` button variant): the text disappears on hover, and the Discord icon is never rendered (always invisible). Both issues occur on light and dark mode. ## What is the new behavior? - **Studio support sidebar**: force a white background on all colour schemes via `bg-white hover:bg-white/90`, matching the existing pattern in `DiscordCTACard`. The purple icon and text now render correctly on light mode. - **Shared `secondary` button variant** (`packages/ui`): - The icon container used `text-border-muted`, which blended into the `bg-foreground` button, so the icon was invisible. Changed to `text-background` to match the button's text colour. - The hover state used `text-border-stronger`, which had no contrast against the background, so text disappeared on hover. Changed to `text-background/80`. - Removed the ineffective `fill="var(--background-default)"` from `IconDiscord` on the `/support` page — the SVG path uses `currentColor`, so the fill prop had no effect. These `secondary` variant fixes apply globally, so other secondary buttons benefit too. | Before | After | | --- | --- | | <img width="868" height="642" alt="CleanShot 2026-07-07 at 15 39 42@2x" src="https://github.com/user-attachments/assets/6897b6ba-311e-43c7-bb5a-7b70017a54cc" /> | <img width="906" height="610" alt="CleanShot 2026-07-07 at 16 21 31@2x" src="https://github.com/user-attachments/assets/16104cf0-8ed1-4857-8467-532c1f7fdb5f" /> | | <img width="636" height="620" alt="CleanShot 2026-07-07 at 15 24 53@2x" src="https://github.com/user-attachments/assets/75f55f6c-eaa7-45ee-94fe-eac513126eff" /> | <img width="656" height="404" alt="CleanShot 2026-07-07 at 16 24 23@2x" src="https://github.com/user-attachments/assets/35649b23-fcdd-42fd-8140-e4aaf394667e" /> | <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit ## Summary by CodeRabbit * **Style** * Refreshed secondary button styling for a more consistent look, including updated hover and loading text/icon colors. * Improved the “Join us on Discord” button’s background and hover appearance. * Simplified the Discord card’s icon rendering to rely on the default icon styling. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Joshen Lim
|
5d0043f59b |
Apply max width for ai assistant panel when maximized (#47974)
## Context Applies a max width for the AI Assistant content area while maximised to improve reading UX <img width="1390" height="958" alt="image" src="https://github.com/user-attachments/assets/296c394d-6bcf-4b0f-b147-ca0253dfa724" /> <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Style** * Improved AI Assistant conversation layout by centering content and limiting its maximum width for easier reading. * Updated support-chat spacing and alignment for a more consistent visual flow. * Refined onboarding content sizing and positioning to improve readability across screen sizes. * Preserved existing chat input behavior while simplifying its styling configuration. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Joshen Lim
|
72d623f1b7 |
Joshen/fe 3775 create a larger assistant workspace for observability (#47954)
## Context Supports maximising the AI assistant to fit the content width which will provide a larger workspace for observability work for example <img width="424" height="179" alt="image" src="https://github.com/user-attachments/assets/f2b84571-0188-4d94-9798-a289496ab543" /> <img width="1388" height="960" alt="image" src="https://github.com/user-attachments/assets/07f52b1f-7b95-4d67-9a5c-151ef036fc7c" /> ## To test - [ ] Can maximize/minimize AI Assistant - [ ] Swapping to another sidebar should bring the panel size back to previous - Only the AI Assistant can be maximized (for now at least) <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added maximize/minimize controls for the AI assistant panel. * Added a keyboard shortcut to quickly maximize the assistant (hidden from settings). * **Improvements** * The main content panel now collapses/resizes automatically based on assistant maximization. * Switching away from the AI assistant now automatically exits maximized mode; returning restores the maximized experience. * Updated the assistant header UI with dynamic label/icon/accessibility and a maximize shortcut when chat isn’t loading. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Danny White
|
805aee289e |
fix(studio): color regressions after theme update (#47794)
## What kind of change does this PR introduce? Bug fix + small shared component ## What is the current behavior? After the recent colour system changes, several UI elements lost contrast in light mode: - `InfoIcon` with a background appeared as a flat grey circle (glyph fill matched the background) - Several buttons forced `text-white`, which no longer contrasts correctly against the updated brand fills - Selected / completed check badges were inconsistent between call sites ## What is the new behavior? - `InfoIcon` uses `text-background` for the glyph fill so the "i" is visible against the grey circle - www primary buttons drop hardcoded `text-white` and use standard `Button` colours - New shared `SuccessCheck` in `ui` for selected state and completion progress (green circle; white check in light mode, black check in dark mode) - Documented in the design system with selected + progress examples Note on `InfoIcon`: [#47933](https://github.com/supabase/supabase/pull/47933) landed a related fix using `text-background-200` (`--card`). This PR keeps `text-background` instead, to match `CheckIcon` / `EyeOffIcon` in the same file and avoid the legacy alias. | Before | After | | --- | --- | | <img width="688" height="268" alt="CleanShot 2026-07-10 at 11 30 25@2x" src="https://github.com/user-attachments/assets/c0276b0c-1023-46c8-805c-35a22def1353" /> | <img width="664" height="278" alt="CleanShot 2026-07-10 at 11 29 43@2x" src="https://github.com/user-attachments/assets/3508cf37-1b48-4fb7-a939-83522feb44f1" /> | | <img width="468" height="550" alt="CleanShot 2026-07-10 at 11 31 04@2x" src="https://github.com/user-attachments/assets/acd42273-15ce-4fb2-9d0c-5a43ac23073c" /> | <img width="460" height="540" alt="CleanShot 2026-07-10 at 11 32 50@2x" src="https://github.com/user-attachments/assets/f66fb4bc-81f7-4df1-95f5-63980c4e8537" /> | ## To test Use the staging preview link from this PR and check the following in **light mode** (and spot-check dark mode): **www** - Visit `/404` — "Head back" button should have readable text (not white-on-green) - Visit `/company` — "Join the team" button in the Team section - Visit a product page with a hero CTA (e.g. `/database`, `/realtime`) — primary "Start for free" button **studio** - Database → Replication → deploy a read replica — footer `InfoIcon` next to the pricing line should show a visible "i" inside the grey circle - `/redeem` — select an org; green `SuccessCheck` should match light/dark contrast (white check / black check) - Settings → API → service role key row — red "secret" tag text should be readable **design system** - `/docs/components/success-check` — demo, selected-state, and progress examples |
||
|
kemal.earth
|
cfd6af4be0 |
fix(studio): autoscaling marker positioning (#47958)
## I have read the [CONTRIBUTING.md](https://github.com/supabase/supabase/blob/master/CONTRIBUTING.md) file. YES ## What kind of change does this PR introduce? Small UI fix to stop things bundling into each other on smaller viewports. Autoscaling marker moves from bottom to top to preserve structural integrity. | Before | After | |--------|--------| | <img width="698" height="246" alt="Screenshot 2026-07-15 at 11 28 07" src="https://github.com/user-attachments/assets/9bbb1bb1-a361-4658-9757-30e012098d0c" /> | <img width="687" height="252" alt="Screenshot 2026-07-15 at 11 28 15" src="https://github.com/user-attachments/assets/ca6b6c0a-b511-4136-835f-4321a30765b8" /> | <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Bug Fixes** * Fixed the Autoscaling tooltip placement on the disk space bar so it consistently appears in the correct spot. * Updated the tooltip trigger to use an info icon with a static positioned tooltip, removing the previous animated trigger behavior while preserving existing tooltip content and visibility rules. * Improved tooltip/marker positioning accuracy by recalculating the horizontal placement from the independently-rounded usage segments to prevent drift. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Julian Domke
|
3f2091395e | fix(billing): exit survey does not fire after downgrade via cancellation button (#47956) | ||
|
Jordi Enric
|
e218b24f96 |
fix(studio): disable unified logs on self-hosted (#47727)
## What Disable the Unified Logs feature on self-hosted Studio (`IS_PLATFORM=false`), where there is no backend to support it. Fixes [FE-3747](https://linear.app/supabase/issue/FE-3747). ## Why Unified logs is platform-only. A provider-level guard already forces the `isPlatformOnly` preview flag off on self-hosted, but two gaps let the feature still surface: 1. The logs sidebar renders `<UnifiedLogsBanner />` unconditionally, so self-hosted users saw an **"Enable preview"** button. Clicking it wrote the preview flag to local storage and routed to `/logs`, transiently rendering the unified logs UI. 2. `useUnifiedLogsPreview` derived `isDefaultOptIn` purely from the `unifiedLogsDefaultOptIn` feature flag, with no platform check. ## Changes - `UnifiedLogsBanner` returns `null` when `!IS_PLATFORM`, removing the only entry point to enable the preview on self-hosted. - `useUnifiedLogsPreview` gates both `isEnabled` and `isDefaultOptIn` on `IS_PLATFORM`, making the hook authoritative so no flag/local-storage state can enable unified logs on self-hosted. ## Tests - `FeaturePreviewContext.selfhosted.test.tsx` — the hook never reports unified logs as enabled or default-opt-in on self-hosted, even with the feature flags on. - `UnifiedLogsBanner.selfhosted.test.tsx` — the banner renders nothing on self-hosted. <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Bug Fixes** * Restricted Unified Logs preview and banner availability to supported platform environments. * Prevented unsupported environments from displaying the Unified Logs banner or opting in by default. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Jordi Enric
|
a23f80dd40 |
feat(studio): identify unified logs analytics queries (#47963)
## What Unified Logs fires several `logs.all.otel` requests on load (row list, chart, sidebar facet counts, single-facet counts) plus inspection queries, all with no identifier — indistinguishable in the network tab. Adds a leading `-- unified logs: <what>` SQL comment to each query builder so each request is identifiable at a glance: - row list - severity chart (with bucket function) - sidebar facet counts - single-facet counts (with facet name) - inspect single log by id - edge function console logs for execution ## Notes `--` comments run to end of line; queries are sent multi-line, so the comment doesn't swallow the SQL. <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Diagnostics** * Added descriptive labels to generated log queries, making SQL statements easier to identify in logs and diagnostics. * Added labels for unified log listings, facet counts, sidebar counts, severity charts, individual log inspections, and related console logs. * **Bug Fixes** * No changes to filtering, grouping, query results, or log retrieval behavior. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Mert YEREKAPAN
|
320604d0e7 |
feat(studio): ship both upgrade CTA placements, remove A/B experiment (#47881)
## What Concludes the `upgradeCtaPlacement` experiment ([#45858](https://github.com/supabase/supabase/pull/45858)) by shipping **both** placements permanently and removing the A/B scaffolding. The experiment tested two upgrade-CTA placements against a control: - **`user_dropdown`** — an "Upgrade to Pro" button in the account/user dropdown - **`org_projects_list`** — a plan-usage card in the org project list Both moved paid conversion in the same direction over control with no activation downside, and they live on separate surfaces, so we're keeping both rather than picking one. ## Changes - **Remove the experiment machinery.** Deleted `useUpgradeCtaExperiment` (PostHog `upgradeCtaPlacement` flag read, `control` arm, variant selection, per-org localStorage seeding, exposure tracking) and replaced it with a small `useShowUpgradeCta` hook that gates purely on **free plan + hosted platform**. Both placements now render for every eligible free-plan org. - **Telemetry cleanup.** Removed the `upgrade_cta_placement_experiment_exposed` event (pure experiment scaffolding). Kept `upgrade_cta_clicked` (with its `placement` property) so we can still measure the CTAs going forward; de-experimented its wording. - **Dead code.** Removed the unused `prependCard` prop plumbing added to `ProjectList` / `EmptyStates` (no caller ever passed it — the org card renders via the `<aside>`). - De-experimented the remaining doc comments in `PlanUsageCard`. Gating is unchanged in spirit: paid orgs and self-hosted never see the CTA. `useShowUpgradeCta` waits until the org plan is known before returning true, so the CTA fades in for free users and never flashes for paid users. ## Testing - `pnpm --filter common typecheck` and `pnpm --filter studio typecheck` pass. - eslint + prettier clean on the changed files. - Manually checked <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit - **New Features** - Upgrade/usage CTAs now appear only when the organization is confirmed to be on a free plan, avoiding premature rendering before plan data is available. - Upgrade CTA visibility is now consistently driven across supported project and organization surfaces (including the user dropdown) for org-scoped routes. - **Bug Fixes** - Removed obsolete “prepend” placeholders from the project list loading and grid views to keep card layouts consistent. - **Documentation** - Clarified upgrade CTA telemetry wording and adjusted the frontend telemetry contract. - **Tests** - Added coverage for upgrade-CTA visibility behavior across key scenarios. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Jordi EnricandJoshen Lim
|
92b8ba7cc9 |
fix(studio): give unified logs filter 'Only' button a solid background (#47953)
The "Only" button in the Unified Logs filter facets had no background, so it overlapped the label/count behind it on hover. Added a solid `bg-background`, set the label to 10px, and capitalized it to "Only". Applied to both the sync and async filter checkbox components. ## Before / After Button now sits on an opaque background and no longer bleeds into the row below. <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Style** * Improved checkbox filter hover interactions by hiding count markers when hovering over an option. * Updated the “Only” control styling for clearer visibility, alignment, and hover behavior. * Capitalized the control label to “Only” for improved readability. <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Co-authored-by: Joshen Lim <joshenlimek@gmail.com> |
||
|
|
d23f86021a |
feat(www): Partner Catalog update (#46757)
## Info architecture change around "Partners" The www "integrations" now become more partner-driven. `/partners/integrations` -> now Partner Catalog under `/partners/catalog` (old links redirect to new paths) Moved them close together in the nav dropdown and in the footer <img width="494" height="336" alt="Screenshot 2026-07-09 at 11 06 41" src="https://github.com/user-attachments/assets/a875fef0-0ab8-47ca-8756-d658b27c4892" /> <img width="1149" height="665" alt="Screenshot 2026-07-09 at 11 09 48" src="https://github.com/user-attachments/assets/9631bb72-fe25-4fb4-b1af-9f14a37d02e7" /> ## /partners This page remains untouched in this PR, updates to layout, content and intake form are delegated to #47874 ## /partners/catalog Listed in the [catalog](https://zone-www-dot-com-git-feat-www-partners-pages-supabase.vercel.app/partners/catalog) are now partners. Some partners match with a listing. <img width="1207" height="866" alt="Screenshot 2026-07-09 at 11 14 17" src="https://github.com/user-attachments/assets/b65216be-976f-4ef5-91f8-1ad49da87b45" /> ## /partners/catalog/[partner] Each partner can have one or more "listings" which are either - simple guides - foreign data wrappers - dashboard integrations Integrations available in the dashboard now all have a prominent "Install integration" cta to open it in the dashboard [integrations page](https://supabase.com/dashboard/project/_/integrations). <img width="1269" height="776" alt="Screenshot 2026-07-09 at 11 16 51" src="https://github.com/user-attachments/assets/3c7bb715-ffce-4d0a-905f-9a660c3b1f5a" /> ## Docs Update docs → [Preview](https://docs-git-feat-www-partners-pages-supabase.vercel.app/docs/guides/integrations) - remove "Supabase marketplace" - use "Dashboard Integrations and Partner Catalog - update integrations in sidenav to link to updated /partners/catalog/** listings <img width="1520" height="696" alt="Screenshot 2026-07-15 at 12 54 47" src="https://github.com/user-attachments/assets/9f5a2794-4536-4299-97df-9732d3d75b4c" /> <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added a Partner Catalog experience with search, category filters, official-partner toggle, responsive filtering (sidebar + bottom sheet), grid/list views, and featured partners. * Added Partner Catalog detail pages with tabbed listings, MDX-rendered content, image gallery with zoom overlay, and “add/install” actions. * **Improvements** * Updated “Become a Partner” layout and form support for prefilled values and checkbox-group fields (including validation). * Updated navigation/footer/docs and partner tile links to use Partner Catalog routes; expanded redirects from legacy integrations paths. * Added public agent-skills discovery manifest. <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Co-authored-by: Alan Daniel <stylesshjs@gmail.com> Co-authored-by: Alex Hall <alex.hall@supabase.io> Co-authored-by: Miranda Limonczenko <miranda.limonczenko@supabase.io> |
||
|
Jordi Enric
|
b02ad91319 |
fix(studio): stop unified logs bar chart animation replay on click FE-3912 (#47910)
## Problem In the unified logs dashboard, clicking an activity bar in the chart restarted the fly-by entrance animation instead of selecting that bar. There was also no way to select a single bar by clicking it (only drag-to-select worked), and the selection menu was centered over the selection rather than anchored to its start. ## Fix - Disabled Recharts' `isAnimationActive` on the stacked `Bar` elements so re-renders from selection state no longer replay the entrance animation. - Clicking a single bar now selects that bar's full time bucket and opens the "Filter logs to selected range" menu, the same as dragging across it. - The menu is anchored to the start (leftmost pixel) of the selection instead of the mouse release position. ## How to test - Open a project's unified logs page - Wait for the activity chart to finish its initial load - Click on a single bar - Expected result: the bar is highlighted, the chart does not replay its entrance animation, and the range-filter menu appears anchored at the start of that bar - Drag across multiple bars - Expected result: the range-filter menu appears anchored at the start of the dragged selection, not centered over it <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit - **Bug Fixes** - Fixed chart highlight popovers to anchor consistently to the start of the selected range. - Improved popover behavior so it updates correctly when the position changes. - Corrected zoom-in filtering when the selected range contains a single timestamp. - **UI Improvements** - Reduced conflicts between tooltips and selection popovers by rendering tooltip content only when appropriate. - Disabled stacked-bar animations for error/warning/success to make chart interactions feel steadier. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Joshen Lim
|
66691ad9ce |
Joshen/fe 3384 prevent ai assistant chat panel from closing after query (#47951)
## Context 2 problems that this PR addresses - but these are in general due to the mobile UI behaviour - Sidebars (e.g Help Panel, Advisor Panel, etc) closes whenever there's a route change - This is happening because of `StudioMobileSheetNav`'s `handleOpenChange` interfering with the sidebar visible state - Am opting to not render `StudioMobileSheetNav` at all unless on mobile - On mobile, if you're on the Advisor Panel, clicking the "Menu" button closes the mobile sheet - Clicking on the Menu button seems to be triggering `MobileSheetNav`'s `onOpenChange` - My suspicion is because of state asymmetry between two independent stores (MobileSheetContext and SidebarManagerState - these 2 are a bit too complex imo) - Am opting to skip calling `onOpenChange` in `StudioMobileSheetNav` if the click target is within the Floating toolbar ## To test - [x] Desktop: Have the assistant panel open and change routes, panel should stay open - [x] Mobile: Open the assistant panel, then switch to the menu, panel should stay open <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Bug Fixes** * Improved mobile navigation interactions so the navigation sheet remains open while toolbar actions are being selected. * Mobile navigation now displays only on smaller screens, preventing it from appearing at medium and larger breakpoints. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Riccardo Busetti
|
c9eb33abc7 | feat(pipelines): Add new cost estimation dialog (#47915) | ||
|
Joshen Lim
|
c97bc6282d |
Adjust AIAssistantHeader (#47912)
## Context As per PR title, just adjusting the Assistant's header a little to improve the UX ### Before <img width="442" height="71" alt="image" src="https://github.com/user-attachments/assets/c714e264-7724-451a-aeaf-7ced456d0639" /> ### After <img width="438" height="77" alt="image" src="https://github.com/user-attachments/assets/11467bf2-7306-4ec9-80e0-2aadd959eff1" /> ## Changes involved - Shift permission settings into "More" dropdown - Chat selection is now "history" - Added keyboard shortcuts for history and copy chat ID <img width="185" height="95" alt="image" src="https://github.com/user-attachments/assets/fc1c9bdc-9180-48ba-940f-2f39fef53a1d" /> <img width="287" height="159" alt="image" src="https://github.com/user-attachments/assets/8f597306-9ea0-4282-884b-722bab16e4d0" /> - Chat name is now clickable to directly edit it - Saves on Enter or on blur - Resets on Esc <img width="433" height="70" alt="image" src="https://github.com/user-attachments/assets/f0c6fb4a-c368-4722-97a2-22ae94cc5511" /> <img width="439" height="64" alt="image" src="https://github.com/user-attachments/assets/1eb21aaf-3979-433a-acc2-378b47b79e94" /> <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added keyboard shortcuts to toggle the AI Assistant chat history and copy the active chat ID. * Added shortcut hint pills to AI Assistant tooltips and the “More options” menu. * Enabled inline editing of the active chat name with save/cancel and blur support. * **Improvements** * Refreshed AI Assistant header actions and icons (including “New chat” and menu controls) for clearer navigation. * Updated onboarding header styling with an assistant icon/animation. * Standardized shortcut rendering in tooltip pill formatting. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Charis
|
52a25c2ebb |
refactor(sql-editor): extract AI/diff + shortcuts hooks (decompose 5/6) (#47935)
## What Decompose step **5 of 6** for `SQLEditor.tsx`. Extracts the Assistant / diff cluster and the keyboard-shortcut wiring out of the `SQLEditorContent` monolith into two focused, individually-testable hooks: - **`useSqlEditorAi`** — SQL completion (`complete`), the ask-AI prompt flow (`handlePrompt`), accept/discard diff handlers, `onDebug` / `buildDebugPrompt` helpers, `handleDiffEditorMount`, and the fragile diff lifecycle effects (one-shot diff-request drain, diff-editor value sync, ask-AI widget visibility). - **`useSqlEditorShortcuts`** — the registered shortcuts (focus editor, new snippet, format, explain) plus the accept/discard/escape keydown handling. `SQLEditorContent` now composes these hooks alongside the execution/explain hooks landed in decompose 4. ## Behavior-preserving This is a pure extraction. The moved function bodies, effect logic, dependency arrays, and JSX are unchanged from the previous monolith (verified via `git diff` against the pre-decomposition source). In particular: - `useEffectEvent` is preserved for `drainDiffRequest` / `resetDiff`. - `editorMountCount` remains single-owner (passed into the AI hook to drive the one-shot drain). - The untrusted→safe SQL promotion (`acceptUntrustedSql`) continues to happen in the run/explain gesture and warning-modal handlers in `SQLEditorContent`, as close to the explicit user action as possible. The Phase-1 characterization suite (`SQLEditor.test.tsx`, 11 tests) remains green. ## Stack Part of the SQLEditor decomposition stack (1/6 … 6/6). Builds on decompose 4 (execution + explain hooks, #47923). Next: PR6 splits the JSX into panes + final cleanup. <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Improved SQL editor AI assistance, including completion prompts, debugging support, and diff review controls. * Added keyboard shortcuts for accepting or discarding AI-generated SQL changes. * Added shortcuts for focusing the editor, creating snippets, formatting queries, and explaining SQL. * **Bug Fixes** * Prevented SQL execution while reviewing AI-generated differences. * Improved handling of AI diff state during editor loading and interaction. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Jordi Enric
|
c9d7f00b8e |
feat(studio): add Go to API Keys command menu action (#47916)
## Summary - Add a "Go to API Keys" action under the command menu's Navigate section, linking to `/project/[ref]/settings/api-keys` ## Test plan - [ ] Open command menu, search "api key", confirm "Go to API Keys" navigates to the settings page <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added a command-menu option to quickly navigate to the project’s API Keys settings page. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Jordi Enric
|
74c75c6d3a |
fix(studio): cap errors-since-last-deploy log range to 24h DEBUG-173 (#47911)
## Problem The edge function overview's "Errors since last deploy" panel queried ClickHouse from the deploy timestamp to now, with no upper bound. When a function had not been redeployed in a long time, this produced an unbounded query range, which is suspected to have caused a recent uptime incident. ## Fix `getSinceLastDeployLogRange` now clamps the query start to at most 24 hours before now, shared by all three queries this panel issues (invocation list, invocation count, runtime logs). The section title was also updated from "Errors since last deploy" to "Errors in the last 24h" to reflect the new bound. ## How to test - Open an edge function's overview page for a function that was deployed more than 24 hours ago - Confirm the "Errors in the last 24h" panel loads without an excessively large query range - Expected result: the panel only queries the last 24 hours of logs regardless of how old the last deploy was - Run `pnpm test:studio -- EdgeFunctionRecentErrors.utils` and confirm the range-clamping tests pass <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit - **New Features** - Updated the Edge Function errors section to show errors from the last 24 hours. - **Bug Fixes** - Limited error log searches to a maximum 24-hour window, preventing outdated results from appearing. - Improved handling of error time ranges when the last deployment occurred earlier than the available window. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Joshen Lim
|
88f3173920 |
Remove home page skeleton loader, improve base loading behaviour (#47903)
## Context The project's home page has been updated for a while now, but the skeleton loader is still showing the old layout Am opting to remove the skeleton loader entirely and instead improve the loading state of the base UI - adjusts some heights to minimize layout shifts <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Improvements** * Added shimmer loading for project titles while project details are loading. * Refined project connection dropdown behavior: improved URL display/truncation, adjusted dropdown alignment, and more reliable option selection (disabled items remain unavailable). * Streamlined the project page’s loading and transition experience across different project states. * **Tests** * Updated the connect E2E scenario to interact with the Connect sheet via its accessible label before selecting the ORM option. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Alaister YoungandAlaister Young
|
8b82d5c472 |
[FE-3895] fix(studio): fit unified logs table to mobile viewport (#47930)
The unified logs table has a fixed content width of ~1400px, so on mobile it overflowed the viewport — columns were clipped and the header appeared misaligned with the rows (the underlying columns were actually aligned; the table just didn't fit). **Changed:** - Progressively hide the three widest columns on narrow viewports via responsive display classes: `method` from `sm`, `pathname` from `md`, `event message` from `lg`. - On phones only the essential columns remain (checkbox, level, date, log type, status), so the table fits with no horizontal scroll. - Desktop (≥`lg`) is unchanged — all columns render exactly as before. Full row data (method/pathname/event message) is still accessible by clicking a row to open the detail panel. ## To test - Open a project's **Logs** (Unified Logs preview) at a mobile width (~390px), with some API log rows in range. - Confirm the table fits the screen — no horizontal scroll/clipping — and the `DATE` header sits cleanly above the dates. - Widen the browser: `method` appears ~640px, `pathname` ~768px, `event message` ~1024px. - At desktop width, confirm all columns show and header/rows line up as before. - Tap a row on mobile → detail panel opens with the full log (method, pathname, event message). <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Style** * Improved responsive table layouts for unified logs. * Columns now adapt visibility based on screen size, keeping key information accessible on narrow displays. * Event messages flex more naturally when visible. <!-- end of auto-generated comment: release notes by coderabbit.ai --> Co-authored-by: Alaister Young <10985857+alaister@users.noreply.github.com> |
||
|
Cemal KılıçandJoshen Lim
|
e90c91e498 |
fix(studio): show custom OAuth providers in Users table (#45658)
## I have read the [CONTRIBUTING.md](https://github.com/supabase/supabase/blob/master/CONTRIBUTING.md) file. YES ## What kind of change does this PR introduce? show provider info in Authentication -> Users page if user is auth'd with custom providers <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit ## Release Notes * **New Features** * Added comprehensive support for custom authentication providers with appropriate type labeling and visual identification * Improved provider display formatting and icon rendering to properly recognize custom authentication options * Enhanced authentication provider interface operations for better custom provider information handling and management <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Co-authored-by: Joshen Lim <joshenlimek@gmail.com> |
||
|
Charis
|
3fa7f086df |
refactor(sql-editor): extract execution + explain hooks (decompose 4/6) (#47923)
## Summary PR **4 of 6** in the SQLEditor decomposition stack. Extracts the two query-run concerns out of the composition root into cohesive hooks. Behavior-preserving — the characterization suite stays green. ## New hooks - `useSqlEditorExecution` — the execute mutation, the `executeQuery` pipeline (destructive-query gating, lazy title generation, connection resolution), and the `potentialIssues` warning-modal state. - `useSqlEditorExplain` — the execute-explain mutation and the `executeExplainQuery` pipeline. Both pipelines accept an **already-promoted `SafeSqlFragment`**. ## Safe-SQL boundary Per review guidance, `acceptUntrustedSql` promotion stays in the **composition root's** run/explain gesture handlers and warning-modal confirm handlers — as close to the explicit user action as possible, so it's auditable. The hooks never call `acceptUntrustedSql`; they only accept safe SQL. `executeQuery` lists its (now all-stable) dependencies directly rather than suppressing `react-hooks/exhaustive-deps`. ## Verification - `vitest` — 11 characterization tests pass - `pnpm --filter studio typecheck` — clean for SQL editor files - `eslint` / `lint:ratchet` — pass (0 new violations) - `prettier --check` — clean `SQLEditorContent` is down to ~530 lines (from 1056 at the start of the stack). Remaining: PR5 (AI/diff + shortcuts), PR6 (JSX pane split + cleanup). <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Bug Fixes** * Improved SQL query execution safeguards for potentially destructive statements, unsafe updates, and missing row-level security coverage. * Improved EXPLAIN handling, including clearer validation for unsupported multi-statement queries. * Preserved query results, errors, and EXPLAIN output more reliably in the SQL editor. * **Improvements** * Added clearer execution status handling and more consistent editor focus, highlighting, and utility-tab behavior. * Improved connection selection and SQL execution reliability across supported database configurations. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
Jordi EnricandClaude Opus 4.8
|
4096267623 |
feat(api-keys): migrate last-used indicator to ClickHouse endpoint (#47458)
## Problem The "last used" indicator for the legacy `anon` / `service_role` API keys (Project API keys settings) was disabled because it ran a BigQuery `edge_logs` query. It is now re-enabled against the ClickHouse-backed `api_keys.last_used.otel` analytics endpoint. ## Current behavior - The `anon` / `service_role` "last used" indicator is off (the BigQuery-backed query was disabled). ## New behavior - New `useApiKeysLastUsedQuery` hook calls the `api_keys.last_used.otel` endpoint (timestamp params only, no SQL sent), plus its query key and the generated platform API type. - `DisplayApiSettings` reads last-used from this hook instead of posting BigQuery `edge_logs` SQL. The pure `getLastUsedAPIKeys` shaper is kept and unit-tested. Still gated by the `showApiKeysLastUsed` flag. - Removed the disabled secret-keys (`sb_secret_`) BigQuery last-used path, which has no ClickHouse endpoint to migrate to: drops the dead `useLastSeen` query, the `APIKeyRow` "Last Used" column, and the unused `showLastSeen` prop. - Reworded the delete-confirmation copy to be accurate for both secret and publishable keys. ## Additional context - Backed by the platform endpoint in supabase/platform#34892 (merged and deployed). - Scope: `anon` / `service_role` legacy keys. Secret/publishable and JWT signing-key "last used" are follow-ups, pending the endpoint returning those key types. <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **Improvements** * Updated API key settings to show “last used” activity for the past 24 hours using a dedicated data source and time window. * Added clearer messaging when recent API key activity fails to load. * Removed the “Last Used” column from API key management tables. * **Bug Fixes** * Improved mapping so “last used” values correctly match the intended key and role. * Updated API key deletion confirmation to explain required backend changes and resulting unauthorized behavior. * **Tests** * Added unit tests to validate “last used” computation and edge-case filtering. <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com> |
||
|
Andrew ValleteauandClaude Fable 5
|
768ea1001b |
fix(studio): scope table editor introspection CTEs to target table OID (#47894)
## I have read the [CONTRIBUTING.md](https://github.com/supabase/supabase/blob/master/CONTRIBUTING.md) file. YES ## What kind of change does this PR introduce? Bug fix (performance), plus a regression-guard test suite and docs. ## What is the current behavior? Studio's introspection queries in `@supabase/pg-meta` do `O(catalog)` work for per-table requests. On databases with very large catalogs (hundreds of thousands of relations/constraints — real deployments reach this) they take tens of seconds per dashboard interaction, trip `statement_timeout`, and create heavy CPU/memory pressure when several tabs open concurrently. Two instances of the same bug class: **1. Table Editor query (`getTableEditorSql`)** — fetches metadata for ONE table by OID, but five catalog scans are unscoped and only filtered at the top-level join: - `primary_keys` CTE — scans all of `pg_index` (`where i.indisprimary`) - `index_cols` CTE — scans all unique indexes - `relationships` CTE — scans every FK in `pg_constraint` (and is scanned twice by the two subplans) - `uniques` subquery (inside `columns`) — scans all single-column unique constraints - `check_constraints` subquery (inside `columns`) — scans all single-column check constraints The planner cannot push the outer join qual into grouped / `distinct on` subqueries, so each is computed over the full catalog and thrown away. `tables-paginated.ts` was previously rewritten to avoid exactly this pattern; the single-table query never got the same treatment. **2. Entity definitions (`getTableDefinitionSql` / `getEntityDefinitionsSql`)** — the vendored `pg_get_tabledef` plpgsql function scans the entire `information_schema.columns` view once **per column** (plus `information_schema.tables` once per call) just to decide whether a name needs double-quoting — a pure string property of a name it already holds — and its per-index partial-index lookup casts `relnamespace::regnamespace::text` across every `pg_class` row. On a 12K-table catalog this makes a single entity's DDL cost ~3.7s and a default 100-entity definitions page ~6 minutes. ## What is the new behavior? **Fix 1 — scope the Table Editor CTEs to the requested OID** (`id` is validated non-null and interpolated via `literal()`, same as the existing `base_table_info` filter): - `primary_keys` / `index_cols`: `and i.indrelid = <id>` - `relationships`: `and (c.conrelid = <id> or c.confrelid = <id>)` - `uniques` / `check_constraints`: `and conrelid = <id>` Semantics are unchanged: the top-level select already filtered every CTE to the target table, so rows for other tables were computed and discarded. The `pg_index`/`pg_constraint` lookups become index scans returning a handful of rows. One residual scan is structural: PostgreSQL has no index on `pg_constraint.confrelid`, so the incoming-FK half of `relationships` is a single filtered seq scan of `pg_constraint` — still one cheap pass instead of materializing every FK row twice. **Fix 2 — remove the O(catalog) scans inside `pg_get_tabledef`**: the information_schema uppercase checks are replaced with direct regex tests on the name in hand (preserving the original's `quote_ident` behavior for schemas that need quoting), and the partial-index lookup is scoped by the already-resolved table OID. Original statements are kept as comments, matching the vendored file's convention. **Regression guard** — so this bug class stays out: - `test/db/stress-catalog.ts` builds a synthetic catalog (default 2,000 tables with PKs, unique + check constraints, FK chains and an FK hub; `PG_META_STRESS_TABLES` scales it to incident size). - `test/db/plan-guard.ts` provides `EXPLAIN (ANALYZE, FORMAT JSON)`-based budget assertions: a query's plan may only seq-scan a scaling catalog if its budget entry carries a written structural justification (e.g. no index on `pg_constraint.confrelid`; no index on `pg_class.relnamespace` for per-schema listings), plus a per-query time bound (the only guard available for opaque plpgsql internals like `pg_get_tabledef`). - `test/sql/studio/catalog-plan-guard.test.ts` applies budgets to the hot-path studio queries: table editor, constraints, FK listing, entity types, tables-paginated, columns, indexes, table/entity definitions, views. Reverting either fix makes the suite fail immediately with the offending scans listed. - `test/sql/studio/table-editor.test.ts` (new — none existed) asserts the Table Editor query's semantics: primary keys, unique indexes, both FK directions, `is_unique`, check definitions, column comments. - A new package `README.md` documents the plan-guard budget entry as a requirement for any new introspection query. ### Validation (synthetic 12,000-table catalog, PostgreSQL 17.6) - **Output equivalence, fix 1:** for 12 relation types (regular, composite PK, partitioned parent + partition, view, materialized view, constraint-free table, FK hub/chain/tail, and a fixture with enums/domains/generated/identity columns and duplicate check constraints), the `entity` jsonb from the old and new query is byte-identical. - **Output equivalence, fix 2:** byte-identical DDL across 13 fixture combinations (serial/identity/generated/array columns, case-sensitive and keyword names, mixed-case schemas, partitions, unlogged + reloptions, partial/expression indexes, external PK/FK/comments/trigger variants). - **Performance, fix 1:** Table Editor query `EXPLAIN ANALYZE` ~1,630ms → ~30ms (~50×); the gap grows with catalog size since the old query is O(catalog) per call. - **Performance, fix 2:** single entity definition 3,672ms → 63ms; a 100-entity definitions page ~6min → 0.87s. The plan-guard bound for `getEntityDefinitionsSql` tightens accordingly from 15s/25 entities to 3s/100 entities (330ms measured at default test scale). Verified locally: `catalog-plan-guard` (12 tests), `table-editor`, `tables-paginated` (16 tests) pass; `typecheck` clean. ### Rollout Per review, the new behavior ships **dark** behind the `pgMetaScopedIntrospection` ConfigCat flag (default off = legacy SQL, kept as full duplicated templates in pg-meta and verified byte-identical to the pre-PR queries). Studio reads the flag in the query hooks and threads it through (flag state is part of the React Query keys). The rollout is staged in the ConfigCat dashboard via user-email targeting (like every other ConfigCat flag): target the reporting user's email first, then a percentage rollout, then 100%. Server-side AI callers of `getEntityDefinitionsSql` stay on the legacy path. Once fully rolled out, delete the legacy templates + flag in a cleanup PR. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Closes: PGMETA-122 <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit - **Bug Fixes** - Improved table editor SQL to correctly scope primary keys, indexes, uniques, checks, and relationships to the selected table. - Optimized table definition SQL to reduce unnecessary catalog scanning for uppercase-name detection and partial-index detection. - **Tests** - Added SQL generator tests for table editor metadata (keys, indexes, relationships, comments, and constraints). - Added catalog query plan guard coverage with a stress catalog and EXPLAIN-based scoping/performance budgets. - **Documentation** - Expanded documentation on catalog query plan safeguards and how to keep new introspection queries properly scoped. <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
