move branches, but make it *graphic* by estib-vega · Pull Request #12696 · gitbutlerapp/gitbutler

estib-vega · 2026-03-05T13:56:03Z

Implement the new way of moving branches by using the graph and new rebase engine.

Add mutation primitives for disconnecting and inserting segments in the graph
Add an API that enables moving branches using the graph (and mark the previous version as legacy)
Add a but command for moving branches

Usage of the but command

but branch move [branch-to-move] [target-branch]
# or 
but stack [brach-to-move] [target-branch]

Tests

Test at the graph operation level
Test at the move-branch API level
Test at the but CLI level

fixes GB-1089

vercel · 2026-03-05T13:56:07Z

The latest updates on your projects. Learn more about Vercel for GitHub.

1 Skipped Deployment

Project	Deployment	Actions	Updated (UTC)
gitbutler-web	Ignored	Preview	Mar 9, 2026 2:18pm

Copilot

Pull request overview

This PR introduces a new “move branch” implementation that operates on the workspace graph (via the new rebase engine), exposes it through a new API surface, and adds a but branch move CLI command while renaming the previous move-branch endpoint/command as legacy.

Changes:

Added graph-based branch moving in but-workspace and exposed it through but-api and but CLI.
Extended graph rebase mutation primitives (disconnect now supports optionally skipping reconnection) to enable segment moves.
Renamed existing UI/backend move-branch wiring to *_legacy and added fixtures/tests for the new behavior.

Reviewed changes

Copilot reviewed 24 out of 24 changed files in this pull request and generated 12 comments.

Show a summary per file

File	Description
crates/gitbutler-tauri/src/main.rs	Switches Tauri invoke command from `move_branch` to `move_branch_legacy`.
crates/but/tests/fixtures/scenario/two-stacks-one-single-and-ready-to-mingle-one-double.sh	Adds a CLI test scenario fixture with two stacks to exercise branch moves.
crates/but/tests/but/command/branch/move_branch.rs	Adds CLI-level tests for `but branch move` (by name and CLI IDs).
crates/but/tests/but/command/branch/mod.rs	Registers the new `move_branch` test module.
crates/but/src/utils/metrics.rs	Adds metrics mapping for the new `branch move` subcommand.
crates/but/src/lib.rs	Wires `branch move` into CLI dispatch; refactors branch subcommand handling.
crates/but/src/command/mod.rs	Makes `command::branch` module always available (not only non-legacy builds).
crates/but/src/command/legacy/branch/mod.rs	Refactors legacy branch command handling into individual functions.
crates/but/src/command/branch/move_branch.rs	Implements the CLI operation that calls workspace graph move + materialization.
crates/but/src/command/branch/mod.rs	Exposes `move_branch` (and keeps `apply` behind non-legacy feature).
crates/but/src/args/metrics.rs	Adds `BranchMove` as a metrics command name.
crates/but/src/args/branch.rs	Adds the `branch move` clap subcommand and arguments.
crates/but-workspace/tests/workspace/branch/move_branch.rs	Adds workspace-level tests covering stack/segment move scenarios.
crates/but-workspace/tests/workspace/branch/mod.rs	Registers the new workspace move-branch tests module.
crates/but-workspace/tests/fixtures/scenario/ws-ref-ws-commit-single-stack-double-stack.sh	Adds a workspace fixture scenario with two stacks.
crates/but-workspace/src/branch/move_branch.rs	Adds the core graph-based branch-move implementation.
crates/but-workspace/src/branch/mod.rs	Exports the new `move_branch` API from the workspace branch module.
crates/but-server/src/lib.rs	Renames server route from `/move_branch` to `/move_branch_legacy`.
crates/but-rebase/tests/rebase/graph_rebase/disconnect.rs	Updates disconnect tests for the new `disconnect_segment_from(..., skip_reconnect_step)` signature.
crates/but-rebase/src/graph_rebase/mutate.rs	Extends disconnect primitive with `skip_reconnect_step` and adjusts reconnect logic.
crates/but-graph/src/api.rs	Adds `Graph::find_segment` helper for looking up segments by index.
crates/but-api/src/legacy/virtual_branches.rs	Renames legacy `move_branch` API to `move_branch_legacy`.
crates/but-api/src/branch.rs	Adds a new graph-based `move_branch` API.
apps/desktop/src/lib/stacks/stackService.svelte.ts	Updates desktop client to call the renamed legacy move-branch command.

Copilot

Pull request overview

Copilot reviewed 24 out of 24 changed files in this pull request and generated 3 comments.

Copilot · 2026-03-05T14:38:47Z

+        Subcommands::Branch(branch::Platform { cmd }) => match cmd {
+            #[cfg(not(feature = "legacy"))]
+            None => todo!("implement list and call recursively"),
+            #[cfg(feature = "legacy")]
+            None => {
+                let mut ctx = setup::init_ctx(
+                    &args,
+                    InitCtxOptions {
+                        background_sync: BackgroundSync::Enabled,
+                        ..Default::default()
+                    },
+                    out,
+                )?;
+                command::legacy::branch::handle_no_subcommand(&mut ctx, out)
            }
-        }
+            #[cfg(feature = "legacy")]
+            Some(branch::Subcommands::List {
+                filter,
+                local,
+                remote,
+                all,
+                no_ahead,
+                review,
+                no_check,
+                empty,
+            }) => {
+                let mut ctx = setup::init_ctx(
+                    &args,
+                    InitCtxOptions {
+                        background_sync: BackgroundSync::Enabled,
+                        ..Default::default()
+                    },
+                    out,
+                )?;
+                command::legacy::branch::list_branches(
+                    &mut ctx, out, filter, local, remote, all, no_ahead, review, no_check, empty,
+                )
+            }


Regression: the legacy but branch ... paths no longer call .emit_metrics(metrics_ctx) (previously the legacy branch handler did). This will stop emitting telemetry for branch list/show/new/delete and the no-subcommand default; consider wrapping these arms with .emit_metrics(metrics_ctx) (or emitting once after the match).

Copilot · 2026-03-05T14:38:48Z

+
+use crate::{CliId, IdMap, id::parser::parse_sources, utils::OutputChannel};
+
+/// Move a branchon top of another


Typo in the doc comment: branchon → branch on (missing space).

Suggested change

/// Move a branchon top of another

/// Move a branch on top of another

Copilot · 2026-03-05T14:38:48Z

+    pub fn find_segment(&self, segment_idx: SegmentIndex) -> Option<Segment> {
+        if self.inner.contains_node(segment_idx) {
+            Some(self.inner[segment_idx].clone())
+        } else {
+            None
+        }


find_segment() clones and returns an owned Segment, which can be unnecessarily expensive (segments can contain many commits). Consider returning Option<&Segment> (e.g. via self.inner.node_weight(segment_idx)) and let callers clone only what they need.

Suggested change

pub fn find_segment(&self, segment_idx: SegmentIndex) -> Option<Segment> {

if self.inner.contains_node(segment_idx) {

Some(self.inner[segment_idx].clone())

} else {

None

}

pub fn find_segment(&self, segment_idx: SegmentIndex) -> Option<&Segment> {

self.inner.node_weight(segment_idx)

Copilot

Pull request overview

Copilot reviewed 24 out of 24 changed files in this pull request and generated 2 comments.

Copilot · 2026-03-06T06:34:00Z

+    let branch_ref_name_str = &format!("refs/heads/{branch_name}");
+    let target_ref_name_str = &format!("refs/heads/{target_branch_name}");
+
+    let but_workspace::branch::move_branch::Outcome { rebase } =
+        but_workspace::branch::move_branch(
+            &ws,
+            editor,
+            branch_ref_name_str.try_into()?,
+            target_ref_name_str.try_into()?,


let branch_ref_name_str = &format!(...) (and the same for target_ref_name_str) borrows a temporary String, which won’t live long enough. This should be an owned String (or build the FullNameRef directly) and then pass a stable &str/&String into try_into().

Suggested change

let branch_ref_name_str = &format!("refs/heads/{branch_name}");

let target_ref_name_str = &format!("refs/heads/{target_branch_name}");

let but_workspace::branch::move_branch::Outcome { rebase } =

but_workspace::branch::move_branch(

&ws,

editor,

branch_ref_name_str.try_into()?,

target_ref_name_str.try_into()?,

let branch_ref_name = format!("refs/heads/{branch_name}");

let target_ref_name = format!("refs/heads/{target_branch_name}");

let but_workspace::branch::move_branch::Outcome { rebase } =

but_workspace::branch::move_branch(

&ws,

editor,

branch_ref_name.as_str().try_into()?,

target_ref_name.as_str().try_into()?,

Copilot · 2026-03-06T06:34:01Z

+        /// The branch being moved.
+        #[clap(value_name = "BRANCH")]
+        branch: String,
+        /// The branch to move the other on top off.


Typo in help text: “on top off” → “on top of”. This text shows up in --help, so it’s worth correcting.

Suggested change

/// The branch to move the other on top off.

/// The branch to move the other on top of.

Byron

I am loving how working with the new workspace and rebase engine can yield new mutations quickly!

With that said, I think we should have some tests to show what happens with purely virtual branches, just because I think that with meta this probably won't yield expected results. It's fine, but maybe we can just limit the participating branches to those that aren't empty.

Lastly, I set this PR back to draft as I think the stringy types in but-api must be fixed, so we can avoid falling back to former patterns of an API that is dominated by its JSON nature.

Finally, let me state that I spent only a little time with this, and didn't review anything outside of the api and the workspace, nor did I look at how this mutation is actually done or its logic. I did skim the tests to see what it can do though, and found them very easy to follow 🩷.

Byron · 2026-03-06T06:17:57Z

+        .segments
+        .iter()
+        .enumerate()


I think this is iter().position().

Will fix 👍

Byron · 2026-03-06T06:27:26Z

+        .base_segment_id
+        .and_then(|segment_idx| workspace.graph.find_segment(segment_idx));
+
+    let parents_to_disconnect = match (stack_base_segment, graph_base_segment) {


This wants to be stack_base_segment.or(graph_base_segment) to avoid the duplication below.

They'r different types, so I won't let me do that :/

Segment and StackSegment

Copilot

Pull request overview

Copilot reviewed 26 out of 26 changed files in this pull request and generated 6 comments.

Copilot

Pull request overview

Copilot reviewed 26 out of 26 changed files in this pull request and generated 2 comments.

Copilot · 2026-03-09T12:22:42Z

+    for stack_entry in &stacks {
+        if stack_entry.heads.iter().all(|b| b.name != *branch_name) {
+            // Not found in this stack,
+            continue;
        }
-        Some(Subcommands::Show {
-            branch_id,
-            review,
-            files,
-            ai,
-            check,
-        }) => {
-            show::show(ctx, &branch_id, out, review, files, ai, check)?;
-            Ok(())
+


In delete(), this branch lookup compares a BString (b.name) against *branch_name (a String). This won’t compile and also mixes encodings (bytes vs UTF-8). Compare using bytes (e.g., b.name.as_ref() == branch_name.as_bytes()) or convert b.name with to_str() and compare to branch_name (handling non-UTF8 names appropriately).

Copilot · 2026-03-09T12:22:43Z

+
+        let mut ws_meta = workspace.metadata.clone();
+
+        let (source_stack, subject_segment) = source;
+        let (_, target_segment) = destination;


ws_meta is initialized as workspace.metadata.clone() and returned even when it isn’t modified (the only mutation is inside the “empty branch” conditional). Downstream callers treat Some(ws_meta) as “needs to be persisted”, so this can cause redundant metadata writes on every non-empty move. Consider returning None unless the metadata actually changes (e.g., start with let mut ws_meta = None; and only clone+mutate when needed), or update the doc/field name to reflect that it may always mirror existing metadata.

Yeah, this probably means it should be set to None when it's not changed.

Byron

Thanks a lot!

I have pushed a refactor commit which straightens out the use of RefMetadata and further limits the move-branch functions to only work if it's a traditional managed commits. For everything else one would definitely need tests.

However, it's probably fair to kick that can hard and say that once meta is backed by the DB officially, and but-rebase alters it, move-branch will naturally work in these cases. So no need to put time into it beyond what 95% of the users need right now.

I did not look at:

tests
integration into but-api or but-rebase
basically anything except for but-workspace, and even there I only dealt with typing/superficial things.

Codex found two issues, one of which is related to a missing edit-mode check. The other one is about possibly missing metrics, both seem worth another look.

[P1] Dispatch legacy branch moves through the guarded path

Repro:

Build/run the default but binary (legacy is enabled by default in crates/but/Cargo.toml).
In a repo with applied GitButler stacks, run but setup.
Leave open-workspace mode, e.g. git checkout main.
Run but branch move A C.

Expected:
This should fail through the legacy guarded path, the same way other legacy branch mutations do, because gitbutler_branch_actions::move_branch enforces ensure_open_workspace_mode(...) and records a MoveBranch snapshot for undo.

Actual:
but branch move now dispatches directly to command::branch::move_branch from crates/but/src/lib.rs, so it bypasses the legacy wrapper. In edit/outside-workspace mode it can still rewrite refs/workspace metadata, and the move is no longer recorded in the oplog.

[P2] Re-wrap branch dispatch in metrics emission

Repro:

Compare the old and new Subcommands::Branch(...) dispatch in crates/but/src/lib.rs.
Notice that match_subcommand() still creates metrics_ctx.
Notice that the new Subcommands::Stack { .. } arm and the refactored Subcommands::Branch(...) arm now return raw Result<()> values instead of calling .emit_metrics(metrics_ctx).

Expected:
but branch ... and but stack ... should emit the same duration/error telemetry as other CLI commands, including the new CommandName::BranchMove mapping added in this patch.

Actual:
No metrics emission happens for these paths anymore, so branch commands silently stop reporting telemetry. That means the newly added BranchMove metric will never fire.

Byron

And some notes on the ominous refactor commit. I made it because I need to test what I feel is right IRL to know it's right.
Also, I get that it might work better for you with comments, so here they are, and it should be fine to drop the commit and redo it based on the gist.

Byron · 2026-03-09T12:29:16Z

 ///
 /// Returned by [function::move_branch()].
 #[derive(Debug)]
-pub struct Outcome<WorkspaceMeta> {


This is a strong smell, and I noticed that LLMs have a hard time with RefMetadata when I tried just now. Basically everything it comes up with is probably not what you want.

I don't like this default assumption that LLMs came up with my changes.

This I did on my own. Would you mind explaining why this is a smell, so that I can avoid that next time?

I don't like this default assumption that LLMs came up with my changes.

Yeah, I will avoid mentioning it. Judging from myself, all I get to do is reviews and refactors these days. And since I am a late-bloomer, I just assume everyone else uses it more :D.

This I did on my own. Would you mind explaining why this is a smell, so that I can avoid that next time?

It makes no sense to use a generic type for a struct for which only one implementation exists. But it doesn't look that way when looking at this Handle type and I really hope that it can be simplified. Something along these lines should be possible once it's finally in the database, at which point there isn't even a need for RefMetadata as trait anymore.

So it's a deal, the trait will go away.

Yeah, the motivation here was to maintain dynamism with the correct types. I assumed that if we're dealing with the implementation of a trait, then it's feasible to get different workspace handles as well.

Didn't know there was only the one

Byron · 2026-03-09T12:30:33Z

+    /// It should replace the actual workspace metadata to configure moved 'virtual' branches segments, if `Some()`.
+    pub ws_meta: Option<but_core::ref_metadata::Workspace>,


It's still quite undefined what should happen without metadata, but it seems fair to just ignore it as there is nothing we can do for now (unless one day there is a case for changing this, but I doubt it).

Byron · 2026-03-09T12:31:18Z

+    pub fn move_branch(
        workspace: &but_graph::projection::Workspace,
        mut editor: Editor,
-        meta: &mut M,


The key here is that but-graph already collects metadata, and thus it should be queried from the Workspace.
This makes handling the metadata update here quite straightforward, but forces the caller to set it if present. Fair enough.

Byron · 2026-03-09T12:32:28Z

+            WorkspaceKind::ManagedMissingWorkspaceCommit { .. } => {
+                bail!("Moving branches currently need a workspace commit")


Let's be even more conservative here, and opt it in with a barrage of tests.
Maybe that won't be needed if the rebase engine takes care of this, so postponing complexity seems like the way to go.

Byron · 2026-03-09T12:33:03Z

+
+        let mut ws_meta = workspace.metadata.clone();
+
+        let (source_stack, subject_segment) = source;
+        let (_, target_segment) = destination;


Yeah, this probably means it should be set to None when it's not changed.

Byron · 2026-03-09T12:34:23Z

        .base_segment_id
        .map(|segment_idx| &workspace.graph[segment_idx]);

-    let parents_to_disconnect = match (stack_base_segment, graph_base_segment) {


I replaced this as it wasn't a match to me. Please double-check, I was lazy and asked my bot to do it 😅.

Gotcha. I guess this is purely stylistic, though.

Happy to keep talking about this offline some other time.

I do think if … else is less ambiguous here and I'd prefer it unless exhaustiveness is needed. Using match won't kill me though, it just confuses me.

Makes sense to me. If else is easier to read 👍

Byron

There are still a couple of legit-looking LLM comments to look into, but besides that I think it's good to go.
Again with the note that I didn't actually try using it.

estib-vega · 2026-03-09T12:50:40Z

+fn set_workspace_metadata(
+    meta: &mut impl RefMetadata,
+    ws: &but_graph::projection::Workspace,
+    ws_meta: Option<but_core::ref_metadata::Workspace>,
+) -> anyhow::Result<()> {
+    if let Some((ws_meta, ref_name)) = ws_meta.zip(ws.ref_name()) {
+        let mut md = meta.workspace(ref_name)?;
+        *md = ws_meta;
+        meta.set_workspace(&md)?;
+    }
+    Ok(())
+}


@Byron Curious, why is this necessary?

Yeah, right!

Git + metadata = graph -> workspace. Then move copies the workspace metadata, and alters it. For refreshes of the workspace it wants meta to re-read reference metadata. So meta has to get the new metadata which is associated with a reference.
It is implicit knowledge that ws.ref_name() is the gitbutler/workspace typically, and that's where we want to set the changed metadata in the metadata store (vb.toml + db mirror right now).

*md = ws_meta;

This is the interesting line as one can't set the data directly. This is me predicting the future, thinking it will be easier when it's an actual database.

But probably it's just helping with the setup now, and in future all this will go away and be a very dumb assignment.

Gotcha. Thanks!

Use the newly implemented graph mutations to move branches Prefer direct segment access in `but_graph::Graph` After all, we may assume that segment ids exist in the graph. And even if not, we can access the inner petgraph directly thanks to a DeRef implementation.

The original `move_branch` API is not marked as legacy.

graphically as in using graph operations. Add a new non-legacy API to move branches around. Show a more idiomatic usage of the API for `move_branch`

Allow for moving branches on top of other branches in the CLI

Add tests for the move branch functionality

Add a top-level command for moving branch and call it stack

- cleanup RefMetadata typing - remove `RefMetadaata` dependency

estib-vega · 2026-03-09T14:27:33Z

Addressed the two open issues found by Codex.

No metrics are emitted in the branch and stack commands
🤦 now we do.
No more oplog entry
Now we do. We snapshot and store if everything went well.

TODO: Do we need to check for open workspace even if we're currently checking for being in a managed workspace?
Not sure. I didn't copy over the ensure_open_workspace call since that's still in the legacy corner.

github-actions Bot added rust Pull requests that update Rust code @gitbutler/desktop CLI The command-line program `but` labels Mar 5, 2026

estib-vega requested a review from Copilot March 5, 2026 13:56

Copilot started reviewing on behalf of estib-vega March 5, 2026 13:56 View session

estib-vega force-pushed the graph-move-branch branch from de7627b to 8d4a964 Compare March 5, 2026 14:05

Copilot AI reviewed Mar 5, 2026

View reviewed changes

estib-vega force-pushed the graph-move-branch branch 2 times, most recently from 5911a34 to 23297a7 Compare March 5, 2026 14:20

estib-vega marked this pull request as ready for review March 5, 2026 14:30

estib-vega requested review from Byron, Caleb-T-Owens and krlvi as code owners March 5, 2026 14:30

Copilot AI review requested due to automatic review settings March 5, 2026 14:30

Copilot started reviewing on behalf of estib-vega March 5, 2026 14:30 View session

Copilot AI reviewed Mar 5, 2026

View reviewed changes

estib-vega force-pushed the graph-move-branch branch from 23297a7 to 9940f96 Compare March 5, 2026 15:05

Copilot AI review requested due to automatic review settings March 6, 2026 06:28

Copilot started reviewing on behalf of Byron March 6, 2026 06:28 View session

Copilot AI reviewed Mar 6, 2026

View reviewed changes

Byron reviewed Mar 6, 2026

View reviewed changes

Byron marked this pull request as draft March 6, 2026 06:46

estib-vega force-pushed the graph-move-branch branch 3 times, most recently from 1a67c15 to 23d1bd9 Compare March 9, 2026 10:22

estib-vega marked this pull request as ready for review March 9, 2026 10:25

Copilot AI review requested due to automatic review settings March 9, 2026 10:25

Copilot started reviewing on behalf of estib-vega March 9, 2026 10:26 View session

Copilot AI reviewed Mar 9, 2026

View reviewed changes

estib-vega force-pushed the graph-move-branch branch from 23d1bd9 to db952e4 Compare March 9, 2026 10:37

Copilot AI review requested due to automatic review settings March 9, 2026 12:16

Copilot started reviewing on behalf of Byron March 9, 2026 12:16 View session

Copilot AI reviewed Mar 9, 2026

View reviewed changes

Byron reviewed Mar 9, 2026

View reviewed changes

Byron approved these changes Mar 9, 2026

View reviewed changes

estib-vega commented Mar 9, 2026

View reviewed changes

estib-vega and others added 7 commits March 9, 2026 15:17

legacy: mark the previous move_branch API

f463b7f

The original `move_branch` API is not marked as legacy.

but api: move branch *graphicallly*

bde1e1b

graphically as in using graph operations. Add a new non-legacy API to move branches around. Show a more idiomatic usage of the API for `move_branch`

but: move branch

dad812a

Allow for moving branches on top of other branches in the CLI

but: test move branch

ad71e08

Add tests for the move branch functionality

but stack

b0b5b1a

Add a top-level command for moving branch and call it stack

refactor

441dce2

- cleanup RefMetadata typing - remove `RefMetadaata` dependency

estib-vega force-pushed the graph-move-branch branch from 9e1c68e to 441dce2 Compare March 9, 2026 14:18

estib-vega merged commit 86e807f into master Mar 9, 2026
36 of 37 checks passed

estib-vega deleted the graph-move-branch branch March 9, 2026 14:42


		use crate::{CliId, IdMap, id::parser::parse_sources, utils::OutputChannel};

		/// Move a branchon top of another

	/// Move a branchon top of another
	/// Move a branch on top of another

	/// The branch to move the other on top off.
	/// The branch to move the other on top of.

		/// It should replace the actual workspace metadata to configure moved 'virtual' branches segments, if `Some()`.
		pub ws_meta: Option<but_core::ref_metadata::Workspace>,

		WorkspaceKind::ManagedMissingWorkspaceCommit { .. } => {
		bail!("Moving branches currently need a workspace commit")

Conversation

estib-vega commented Mar 5, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Usage of the but command

Tests

Uh oh!

vercel Bot commented Mar 5, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 5, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Mar 6, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 6, 2026

Choose a reason for hiding this comment

Uh oh!

Byron left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

estib-vega Mar 6, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

estib-vega commented Mar 5, 2026 •

edited

Loading

vercel Bot commented Mar 5, 2026 •

edited

Loading

Byron left a comment •

edited

Loading

estib-vega Mar 6, 2026 •

edited

Loading