Previously we required AgentSessionInfo all over the place, which meant
there were lots of unnecessary fake ones created all over the place.

Made the methods and functions only take the data they need so we only
use these in history contexts now, as intended.

Release Notes:

- N/A

Making the message editor take all the available space, and improving
the loading state for external agents a bit.

Release Notes:

- N/A

- Make clicking on the project header collapse/expand the group as
opposed to activating the workspace
- Added the "threads" label close to the sidebar toggle button
- Made the open folder icon button open the file finder directly as
opposed to the recent projects popover

Release Notes:

- N/A

Adding some adjustments so the toolbar looks more like the designs we've
been working on. All that changes here are only valid for the v2 feature
flag. Haven't changed anything for today's production version.

Release Notes:

- N/A

Add an explicit `Editor && mode == auto_height` context block. This
ensures that `Shift+Enter` and `Ctrl+Enter` correctly insert a newline
at the cursor position in editors like the AI Agent Panel,
preventing them from inheriting conflicting overrides (e.g., JetBrains
mapping `Shift+Enter` to `editor::NewlineBelow`).

Closes #47269

Release Notes:

- Fixed an issue where `Shift+Enter` would insert a newline at the end
of the text instead of the cursor position in the Agent Panel when using
certain keymaps.

Very much wip

Release Notes:

- N/A

This allows you to now re-use our Zed flake in different side projects
with cargo wrapper script having its path correctly resolved.

Say you have this dir structure:

```
$HOME/dev/zed
$HOME/dev/something
```

Then this now works:

```
$ cd $HOME/dev/something
$ nix develop ../zed
$ cargo version
cargo 1.93.0 (083ac5135 2025-12-15)
```

Release Notes:

- N/A

Closes #49800

Adds `handle_drop` to Item & ItemHandle, which allows an active item in
a pane to consume drop events before the pane does.

Release Notes:

- terminal: Fix drag-and-drop not working in vertical terminal panels

Closes #50631

Before you mark this PR as ready for review, make sure that you have:
- [x] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
- [ ] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)

Release Notes:

- Fixed `formatter: "auto"` silently doing nothing when the first
language server for a language doesn't support formatting (e.g., Dependi
before Tombi for
TOML).

Release Notes:

- Fixed an issue where shared agent thread URLs would not open.

When I am in the middle of editing, pressing Ctrl would
counter-intuitively highlight links even when the mouse cursor is
hidden. This change considers the state of the mouse cursor before
painting links on hover.

Before: Modifier pressed, cursor hidden, link visible

<img width="506" height="518" alt="image"
src="https://github.com/user-attachments/assets/82a59e83-e3cb-490f-b292-148686ec569d"
/>

After: Modifier pressed, cursor hidden (red dot indicates current cursor
position)

<img width="408" height="298" alt="image"
src="https://github.com/user-attachments/assets/c19ed83c-4778-4890-97b9-5155cdcf658b"
/>

Before you mark this PR as ready for review, make sure that you have:
- [x] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
- [x] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)

Release Notes:

- Fixed spurious link highlighting when mouse cursor is hidden

Fixes #50776

Closes #50619

In the conda activation script building procedure, Zed currently
performs a file check for the conda executable on the client side. When
in remote development, this check always fails because the file exists
on the remote host, not the local machine. Since `pet` already handles
existence checks, removing this redundant check allows the activation to
proceed. It is also better to let any potential issues (like
permissions) show up in the terminal rather than silently skipping the
activation.

This addresses the root cause for remote development, which is different
from the approach in #50715 that focuses on shell hooks.

**The video recording**


https://github.com/user-attachments/assets/62967351-e3c5-4814-aec4-b2332940e7e3

Before you mark this PR as ready for review, make sure that you have:
- [x] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
- [x] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)

Release Notes:

- Fixed conda environment not auto-activating in the terminal during
remote development sessions.

Release Notes:

- N/A

---------

Co-authored-by: Eric Holk <eric@zed.dev>

Release Notes:

- N/A

Release Notes:

- Linux: Handle GPU device loss gracefully

Add GPT-5.4 and GPT-5.4-pro as Bring Your Own Key model options for the
OpenAI provider.

**GPT-5.4** (`gpt-5.4`):
- 1,050,000 token context window, 128K max output
- Supports chat completions, images, parallel tool calls
- Default reasoning effort: none

**GPT-5.4-pro** (`gpt-5.4-pro`):
- 1,050,000 token context window, 128K max output
- Responses API only (no chat completions)
- Default reasoning effort: medium (supports medium/high/xhigh)

Also fixes context window sizes for GPT-5 mini and GPT-5 nano (272K →
400K) to match current OpenAI docs.

Closes AI-78

Release Notes:

- Added GPT-5.4 and GPT-5.4-pro as available models when using your own
OpenAI API key.

Fixes a bug introduced in
https://github.com/zed-industries/zed/pull/50850, where we used
incorrect coordinates for the editable range.

Release Notes:

- N/A

I think we might just get it working this time

Release Notes:

- N/A

Release Notes:

- N/A

---------

Co-authored-by: Jakub Konka <kubkon@jakubkonka.com>
Co-authored-by: Oleksiy Syvokon <oleksiy.syvokon@gmail.com>
Co-authored-by: Ben Kunkle <ben@zed.dev>

Closes #ISSUE

Before you mark this PR as ready for review, make sure that you have:
- [ ] Added a solid test coverage and/or screenshots from doing manual
testing
- [ ] Done a self-review taking into account security and performance
aspects
- [ ] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)

Release Notes:

- N/A *or* Added/Fixed/Improved ...

Release Notes:

- N/A

Fixes ZED-5G8

If `DisplayLink::new` fails, `frame_requests` is dropped . It is not
valid to destroy a DispatchSource that is not `resume()`d. So, ensure we
call `resume()` before there's a chance for anything to fail.

Release Notes:

- Fixed a crash that could occur on macOS when changing monitor
configurations

Closes #ISSUE

Before you mark this PR as ready for review, make sure that you have:
- [ ] Added a solid test coverage and/or screenshots from doing manual
testing
- [ ] Done a self-review taking into account security and performance
aspects
- [ ] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)

Release Notes:

- N/A *or* Added/Fixed/Improved ...

Make sure we don't download this file if disable_ai is enabled

Release Notes:

- N/A

Remove the conditional fullscreen path for subagent tool calls so
expanded content consistently renders as a capped preview with the
overlay and fixed height. Also add a pointer cursor to the maximize
control to clarify clickability.

Release Notes:

- N/A

These aren't built in anymore!

Release Notes:

- N/A

Adds five telemetry events to track subagent lifecycle and user
interactions:

- **Subagent Started** — fires on creation or resumption, includes
parent session, subagent session, depth, and whether it was resumed
- **Subagent Completed** — fires when `SpawnAgentTool` finishes,
includes subagent session and status (completed/error)
- **Subagent Toggled** — fires when the user expands or collapses a
subagent card
- **Subagent Stopped** — fires when the user clicks Stop on a running
subagent
- **Subagent Maximized** — fires when the user clicks Maximize to
navigate into a subagent
 
Release Notes:
 - N/A

Closes #ISSUE

Before you mark this PR as ready for review, make sure that you have:
- [ ] Added a solid test coverage and/or screenshots from doing manual
testing
- [ ] Done a self-review taking into account security and performance
aspects
- [ ] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)

Release Notes:

- N/A *or* Added/Fixed/Improved ...

Co-authored-by: Agus <agus@zed.dev>

Before you mark this PR as ready for review, make sure that you have:
- [x] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
- [x] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)

Release Notes:

- N/A

---------

Co-authored-by: Eric <eric@zed.dev>
Co-authored-by: cameron <cameron.studdstreet@gmail.com>
Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
Co-authored-by: Anthony Eid <anthony@zed.dev>
Co-authored-by: John Tur <john-tur@outlook.com>

This PR updates the model-fetching behavior for the Zed provider to wait
until the user is fully signed-in before we attempt to fetch the models.

We now subscribe to the `UserStore` and listen for the
`PrivateUserInfoUpdated` event. If we receive that event and the user is
not currently signed out, then we should be all set to fetch the models.

Closes CLO-424.

Release Notes:

- N/A

Remove the separate `AgentGitWorktreesFeatureFlag` and replace all its
checks with the existing `AgentV2FeatureFlag`. The git worktree
functionality is now gated behind the same flag as agent v2.

Changes:
- Removed `AgentGitWorktreesFeatureFlag` from `feature_flags`
- Replaced all `has_flag::<AgentGitWorktreesFeatureFlag>()` checks with
`has_flag::<AgentV2FeatureFlag>()`
- Simplified toolbar condition that was redundantly checking both flags
- Updated tests accordingly

Closes AI-73

Release Notes:

- N/A

Closes #50627

Before you mark this PR as ready for review, make sure that you have:
- [x] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
- [ ] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)


### Videos:
Before:

https://drive.google.com/file/d/1PahGhfx-wq9cyNqqMlctb9Np7M1Meo71/view?usp=sharing
After:

https://drive.google.com/file/d/135W6MQ9hKBurw5Z7YpHQQhoQwkh1NvUb/view?usp=sharing


Release Notes:

- Fixed clicking on non-selectable picker entries (e.g. section headers)
confirming the currently selected item.

When a diagnostic-triggered prediction returns no edits, don't fall back
to requesting another diagnostic prediction. This was causing an
infinite loop that burned through the API rate limit until hitting 429
errors.


Release Notes:

- N/A

Haiku 4.5 supports extended thinking but not adaptive thinking.

Co-authored-by: Ona <no-reply@ona.com>

When multiple `queue_prediction_refresh` calls land in the same
synchronous frame, all spawned tasks would read the same throttle state
and each send an identical cloud request.

Now the second task would see that the first request has been sent and
would quit.

Release Notes:

- N/A

See these screenshots:

<img width="832" height="564" alt="image"
src="https://github.com/user-attachments/assets/64d04fe7-4be7-4fbf-8536-b4b7a2ea62ed"
/>

<img width="836" height="1094" alt="image"
src="https://github.com/user-attachments/assets/247b215c-8e7e-4eff-8e08-b1b9ee8894b9"
/>

Release Notes:

- Fixed some error messages in OpenAI/Copilot Chat conversations that
were using the Debug representation.

Closes #50170.

Before you mark this PR as ready for review, make sure that you have:
- [x] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
- [x] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)

Release Notes:

- Enable parser options individually to avoid rendering issues

Add some debug logging to the streaming edit file tool, to make it
easier to debug errors.

Can be enabled with 
```
"log": {
  "agent": "debug",
}
```

Release Notes:

- N/A

Fixes a typo in the consent banner check on docs pages, which makes the
opt-out logic work correctly.

Before you mark this PR as ready for review, make sure that you have:
- [x] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
- [x] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)

Release Notes:

- N/A

Nightly only since no tool is using streaming on Preview.

Tools that used streaming could deadlock when the upstream LLM API
stream exited early without ever sending a ToolUse with `input_complete`
set to `true`. This is now fixes by manually dropping the tool input
channels.

Release Notes:

- N/A

Release Notes:

- N/A

Turns out `Editor::newline_below` was not respecting the `read_only`
flag because it uses buffer.edit. `Editor::newline` and
`Editor::newline_above` behaved correctly because they used
`editor.edit` which ignores edits if `read_only` is true.

Added an early return if the editor is read only, also added it to
newline/newline_above for clarity.

Release Notes:

- Fixed an issue where newlines could be inserted into read-only editors

Fixes #50526
Fixes https://github.com/zed-industries/zed/issues/42841
Fixes https://github.com/zed-industries/zed/issues/49875

### Summary

Fix `autosave: on_focus_change` not firing reliably when leaving editors
with nested focus targets (e.g. multibuffer/search flows).

### Root cause

Autosave on focus change was wired to `on_blur` of the item focus
handle.
`on_blur` only fires when that exact handle is the focused leaf, which
misses common descendant-to-outside focus transitions.

## Fix

In `crates/workspace/src/item.rs`, switch autosave subscription from:
     - `cx.on_blur(&self.read(cx).focus_handle(cx), ...)`

to:
     - `cx.on_focus_out(&self.read(cx).focus_handle(cx), ...)`

Autosave behavior and guards remain unchanged:
- only for `AutosaveSetting::OnFocusChange`
- only when focus truly left the item (`!contains_focused`)
- skipped when modal is active (`!has_active_modal`)

### Impact

- Fixes missed autosaves when moving focus from item descendants to
other UI (terminal, file tree, search inputs, etc.).
- No behavior change for other autosave modes (`off`, `after_delay`,
`on_window_change`).

### Video
[Screencast from 2026-03-04
15-23-07.webm](https://github.com/user-attachments/assets/81135999-320f-40f3-9673-7c0460e5a14a)




Release Notes:

- Fixed an issue where `"autosave": "on_focus_change"` would not
reliably work in multibuffers.

---------

Co-authored-by: MrSubidubi <finn@zed.dev>

Release Notes:

- N/A

Closes #47823 

Before you mark this PR as ready for review, make sure that you have:
- [x] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
- [x] No UI changes
 
Release Notes:

- Solves the issue where env variables failed to load when a project
with dir path had `'`(single quote) in them, for example:
`C:\Temp\O'Brien\project_1`.
- added a zed pre-quote for directory paths and zed executable 
- handled pwsh, nushell, cmd, fish, posix, csh, tcsh, rc, xonsh, elvish
based terminals
-  uses `try_quote` for quoting shell paths

Release Notes:

- N/A

Release Notes:

- N/A

Replace the auto-generated `agent-[hash]` branch naming scheme with
names based on historical typewriter models (e.g. `olivetti-a3f9b2c1`,
`selectric-7d2e4f01`).

A static list of 49 typewriter names is used as the pool. When selecting
a name, existing branch names are checked—each branch is split on its
last `-` to extract the typewriter prefix, and any taken prefixes are
excluded from the candidate set. A random name is then picked from the
remaining candidates and suffixed with an 8-character alphanumeric hash.

The name selection logic accepts injected RNG and a list of disallowed
names, making it straightforward to test deterministically with
`#[gpui::test]` and `StdRng`.

Closes AI-71

(No release notes because this is feature-flagged.)

Release Notes:

- N/A