## 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
## What is the current behavior?
When a cell gets moved via `move_cell` operation in
`deriveNotebookDiff`, the `insertedAfter` offset map is not cleared for
that anchor cell. This causes later `insert_cell` operations anchored on
the same (now-moved) cell to apply the stale offset on top of the
correct current-position lookup, resulting in the new cell landing after
the wrong position.
## What is the new behavior?
The offset for an anchor cell is now cleared from `insertedAfter` when
it gets moved, since cells previously inserted after it stay behind at
its old location and should not affect subsequent inserts at its new
position.
A regression test has been added that reproduces the exact ticket
scenario (insert after cell-1, move cell-1 after cell-3, insert after
cell-1 again) and verifies the correct final cell order.
## Additional context
Fixes:
https://linear.app/supabase/issue/FE-4308/insert-anchored-to-a-previously-moved-cell-lands-after-the-wrong-cell
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Fixed notebook cell insertions after moving an anchor cell, ensuring
new inserts appear relative to the anchor’s updated position.
* Preserved the placement of inserts made before the anchor cell was
moved.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## Summary
- Regenerates `packages/api-types` for the content endpoints now that
the Platform API's `notebook` content type has landed (list/get/upsert
`type` enums, plus `UpsertContentBody`'s notebook cell shape with
`_id`/`y_series`). Unrelated schema drift from the same regen
(Warehouse, SSO, notification exceptions, etc.) is excluded — only the
content-endpoint hunks are applied.
- Removes every local widening cast added while the API support was
pending (`content-query.ts`, `content-infinite-query.ts`,
`notebook-query.ts`, `notebook-upsert-mutation.ts`,
`sql-folders-query.ts`).
- What remains is scoped and renamed to match: draft ids
(`generateDraftId`/`isDraftId`), used only for cells created client-side
in the editor before their first save, dropped before they'd ever reach
the backend as a fake `_id`.
## Test plan
- [x] `pnpm typecheck` — clean
- [x] `pnpm --filter studio test` — full suite passes (518 files / 5471
tests)
- [x] `pnpm --filter studio run lint:ratchet` — no new warnings
- [x] `pnpm format` / prettier — clean
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Bug Fixes**
* Improved notebook cell tracking during editing, reordering, insertion,
and deletion.
* Preserved existing cell identifiers while removing temporary draft
identifiers before saving.
* Improved chart configuration for selecting and displaying multiple
Y-axis series.
* Strengthened notebook validation and content persistence behavior.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
---------
Co-authored-by: Joshen Lim <joshenlimek@gmail.com>
## 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?
Refactor, plus one bug fix.
Groundwork for showing the user a preview of what they are approving
when the AI Assistant creates or edits a notebook. No UI in this PR.
Towards FE-4143
## What is the current behavior?
`applyNotebookOperations` resolves an ordered list of notebook
operations into the resulting cells and nothing else. Rendering a diff
for the approval gate needs to know *what happened* to each cell
position, not just where things landed, so there is no way to build the
preview on top of it.
Separately, replacing a cell dropped its id, so `[replace cell-2, insert
after cell-2]` failed with a spurious `unknown_cell_id`.
## What is the new behavior?
`deriveNotebookDiff` resolves operations into one annotated entry per
cell position (`unchanged`, `added`, `removed`, `replaced`, `moved`).
`applyNotebookOperations` becomes a thin projection over its result, so
there is a single interpreter of notebook operations and the diff a user
approves cannot disagree with the cells that get written. The
pre-existing tests pass untouched, which is the evidence that the
projection is faithful.
Notes on the annotations:
- `removed` entries stay in the position the cell used to hold so the
list reads as a diff. This does not perturb insert-anchor arithmetic:
prior inserts still sit contiguously after their anchor.
- Moves that cancel out are downgraded to `unchanged`, since two moves
can anchor on each other and leave every cell where it started. Badging
those as moved would make the preview lie.
- `fromIndex` is the cell's position in the original notebook rather
than in the shifted working order, so `was #3` means what a reader
expects.
A replaced cell now stays addressable as an anchor. Anchoring and
targeting are separate lookups: a replaced cell can be anchored on, but
is never a legitimate target.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Notebook changes now provide a structured view of added, removed,
replaced, moved, and unchanged cells.
* Replaced cells can be used as insertion anchors, while invalid or
duplicate targets are rejected.
* No-op moves are handled as unchanged cells.
* Notebook edits preserve operation ordering and original cell positions
for more predictable results.
* **Bug Fixes**
* Improved notebook operation handling and error reporting for complex
cell edits.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## Summary
- Adds deterministic `list_notebooks`/`get_notebook` fixtures (two
seeded notebooks, each with markdown/database/log cells) and stateful
in-memory `create_notebook`/`update_notebook` mocks to
`apps/studio/lib/ai/tools/mock-tools.ts`, so Braintrust evals can
exercise notebook tool calls without a real project.
- Both write-tool mocks force `needsApproval: false`, matching the
existing `execute_sql`/`deploy_edge_function` mock pattern — the eval
harness filters out tool-parts in `'approval-requested'` state and can
never answer an approval gate.
- All four notebook tools are wrapped from the real `getNotebookTools()`
definitions (only `execute`/`needsApproval` overridden), so evals
validate the model's arguments against the exact production schemas.
- Dedupes `describeOperationError` (previously duplicated between this
new mock and `notebook-tools.ts`) into a single exported
`describeNotebookOperationError` in `notebook-operations.ts`.
**Stacked on #48949** (`feature/notebooks-update-tool`) — this PR's base
branch is that PR, not `master`, because it reuses `update_notebook` and
the shared error helper that only exist there. Merge #48949 first, then
retarget/merge this one.
## Test plan
- [x] \`pnpm --filter studio typecheck\` passes
- [x] \`pnpm --filter studio test\` — all notebook-related suites pass
(\`mock-tools.test.ts\`, \`notebook-tools.test.ts\`,
\`data/content/notebooks/*\`)
- [x] \`eslint\` / \`prettier --check\` clean on all touched files
## Summary
- Pure module (`data/content/notebooks/notebook-operations.ts`) for
applying `update_notebook` cell edits client-side: `insert_cell`
(`after_cell_id` incl. `'start'`), `replace_cell`, `delete_cell`,
`move_cell`.
- Never touches the safe-sql brands — SQL promotion still happens at the
tool-execute boundary, matching `create_notebook`.
- Stacked on #48938. No wiring yet — `update_notebook` tool wiring is
next.
Towards FE-4083
## Test plan
- [x] `pnpm vitest run
data/content/notebooks/notebook-operations.test.ts` — 13 unit tests
covering every op, combinations, and all three error cases.
- [x] `pnpm exec tsc --noEmit` clean
- [x] `pnpm exec eslint` clean on new files
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added support for applying notebook cell operations, including
insertion, replacement, deletion, and movement.
* Operations are applied in a predictable order, with support for
anchoring new cells at the beginning or near existing cells.
* Added validation for invalid references, conflicting operations, and
self-referential moves.
* Added clear handling when operations produce an empty notebook result.
* **Tests**
* Added comprehensive coverage for individual, combined, ordered,
conflicting, invalid, and empty-result notebook operations.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
Charis