Closes #ISSUE

Annotated our `default.json` with `$schema` to get diagnostics, then
fixed the non-language not installed warnings.

Release Notes:

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

Closes #ISSUE

Release Notes:

- settings: Deprecated `code_actions_on_format` in favor of specifying
code actions to run on format inline in the `formatter` array.

Previously, you would configure code actions to run on format like this:

```json
{
  "code_actions_on_format": {
    "source.organizeImports": true,
    "source.fixAll.eslint": true
  }
}
```

This has been migrated to the new format:

```json
{
  "formatter": [
    {
      "code_action": "source.organizeImports"
    },
    {
      "code_action": "source.fixAll.eslint"
    }
  ]
}
```

This change will be automatically migrated for you. If you had an
existing `formatter` setting, the code actions are prepended to your
formatter array (matching the existing behavior). This migration applies
to both global settings and language-specific settings

Closes #ISSUE

Release Notes:

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

Reworded the error message when the `rsync` utility could not be found.

Release Notes:

- N/A

Rendering breaks when both an element and its parent have opacity set.
The following code reproduces the issue:

```rust
struct Repro;

impl Render for Repro {
    fn render(&mut self, _window: &mut Window, _cx: &mut Context<Self>) -> impl IntoElement {
        fn make_box(bg: impl Into<Fill>) -> impl IntoElement {
            div().size_8().bg(bg).hover(|style| style.opacity(0.5))
        }

        div()
            .flex()
            .items_center()
            .justify_center()
            .size(px(500.0))
            .hover(|style| style.opacity(0.5))
            .child(make_box(gpui::red()))
            .child(make_box(gpui::green()))
            .child(make_box(gpui::blue()))
    }
}
```

Before (broken behavior):


https://github.com/user-attachments/assets/2c5c1e31-88b2-4f39-81f8-40060e3fe958

The child element resets its parent and siblings' opacity, which is an
unexpected behavior.

After (fixed behavior):


https://github.com/user-attachments/assets/48527033-b06f-4737-b6c3-0ee3d133f138

Release Notes:

- Fixed an issue where nested opacity is rendered incorrectly.

Release Notes:

- N/A

This PR renames the following actions to make it easier and prioritize
the UI version of interacting with them:

| Before | After |
|--------|--------|
| `OpenSettingsEditor` | `OpenSettings` |
| `OpenSettings` | `OpenSettingsFile` |
| `OpenKeymapEditor` | `OpenKeymap` |
| `OpenKeymap` | `OpenKeymapFile` | 

Release Notes:

- Rename actions to open settings (UI/window and JSON file) as well as
to open the keymap (editor tab and JSON file).

Closes #ISSUE

Release Notes:

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

Refactors zeta2 cli a bit. Merging this by itself to prevent conflicts.

Release Notes:

- N/A

Release Notes:

- N/A

Co-authored-by: Michael Sloan <michael@zed.dev>

Added missing comma after "After installing" and removed duplicated
"that" in developing-extensions.md.

Release Notes:

- N/A

Release Notes:

- N/A

---------

Co-authored-by: John Tur <john-tur@outlook.com>

Closes #ISSUE

Release Notes:

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

Removes the duplication between `gui.rs` doc comments and the `README.md` file.

Release Notes:

- N/A

Closes #ISSUE

Release Notes:

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

This fixes an issue introduced via
[v0.208.0-pre](https://github.com/zed-industries/zed/releases/tag/v0.208.0-pre)
and reported via
https://github.com/zed-industries/zed/issues/34530#issuecomment-3386042577
where, when using fish shell as the default shell and using a Claude
Code thread in Zed, all command were failing because `(command)` in fish
is for command substitution. Using it creates this type of error:

```
fish: command substitutions not allowed in command position. Try var=(your-cmd) $var ...
(npm ci) </dev/null
^~~~~~~~~~~~^
```

or in the editor itself:

<img width="1624" height="1060" alt="image"
src="https://github.com/user-attachments/assets/64fc3126-2cdd-450e-bc85-ef91c56b3705"
/>


Using the appropriate syntax to redirect to stdin for fish fixes the
issue.

Release Notes:

- Fixed redirect stdin command for fish shell

Previously we were always adding a `Navigable` entry for the "new WSL
connection" option in this modal, even though we don't have the
corresponding button on non-Windows. This was causing `menu::SelectNext`
to behave incorrectly (focusing the center pane instead) when `Connect
New Server` was selected on macOS and Linux.

Release Notes:

- Fixed a bug with keyboard navigation in the remote project modal.

Closes #39263

Release Notes:
- N/A 

from
https://github.com/zed-industries/zed/issues/39263#issuecomment-3358220988

> 
> > If you replace that code with
> > 
> > let adapter: IDXGIAdapter1 = unsafe { 
> >    dxgi_factory.EnumAdapters(adapter_index) 
> > }?.cast()?; 
> > 
> > does it not select the right GPU?
>  
> @reflectronic That does seem to select the active gpu for me, meaning
whichever GPU is currently connected. This is a much simpler solution
than the one I have here
(https://github.com/zed-industries/zed/pull/39264 - updated) and while
I'm sure I could imagine someone wanting to choose their GPU to render
Zed on, that may not be something that the application really needs to
support.
> 
> I have a branch with just this as the only change that I can push to
that PR if the simpler solution is preferred.
> 
> ```rust
>         let adapter: IDXGIAdapter1 = unsafe {
>             dxgi_factory.EnumAdapters(adapter_index)?.cast()?
>         };
> ```

Adds docs for Zed hosted Grok models

Release Notes:

- N/A

Gathering LSP declarations in zeta_cli can take a really long time for
big repos and has to be started from scratch if interrupted.

Instead of writing the cache file once we have walked the whole
worktree, we'll now do so incrementally as we complete each file. On
subsequent runs, we'll load as many valid declarations as has been
previously written to the cache, and then continue to request the rest
from the LSP which will append to the existing file as it makes
progress. If the last cache entry is incomplete, we'll truncate the
cache file to the end of the last valid line and continue from there, so
we can just `ctrl-c` without breaking resumability.

Release Notes:

- N/A

- Update `AcpThreadView.insert_selections` to take into account whether
the user is currently editing an existing message and, if it is, insert
the selection into that message instead of the thread's message editor
- Update Window's default keymap to use the `agent::QuoteSelection`
action instead of the deprecated `assistant::QuoteSelection` action
- Introduce `AcpThreadView.active_editor` to allow callers to retrieve
either the thread view's message editor or the editor for the message
being edited, in case `AcpThreadView.editing_message` is not `None`
- Improve `AcpThreadView.focus_handle` to focus on the message being
currently edited in case the user navigates back to the editor and then
to the thread view again, all while editing a message
- Add tests for `AcpThreadView.insert_selections`, ensuring that the
selection is inserted in the message being currently edited, if a
message is being edited, or the thread view's message editor if no
message is being edited

Closes #39693 

Release Notes:

- Improved `agent: quote selection` to also work for a message that was
already sent but is being edited

---------

Co-authored-by: Ben Brandt <benjamin.j.brandt@gmail.com>

The feedback modal did not match our keyboard-driven design. We can
revisit this later if we want, but for now, removing it makes sense. All
actions have been inlined in the `Help` menu to maintain
discoverability.

Additionally, not all feedback-based actions in the command palette were
namespaced under `feedback:`, and now they are, so they can all be found
there easily.

Release Notes:

- Notice: The `Give Feedback` modal has been removed. The options to
file bug reports, feature requests, email us, and open the Zed
repository can now be found within the `Help` menu directly. The command
palette actions have undergone the following changes:

- `feedback: give feedback` (removed)
- `feedback: file bug report` (no change)
- `zed: request feature` → `feedback: request feature`
- `zed: email zed` → `feedback: email zed`
- `zed: open zed repo` → `contribute: open zed repo`

Previously we were guessing the context window size here:
https://github.com/zed-industries/zed/blob/8c3f09e31e3588a2494042dfe77bb3dccab1f7ba/crates/ollama/src/ollama.rs#L22

This is inaccurate and must be updated manually. This PR ensures that we
extract the context window size from the request in the same way that
the Ollama CLI does when running `ollama show <model-name>` (Relevant
code is
[here](https://github.com/ollama/ollama/blob/3d32249c749c6f77c1dc8a7cb55ae74fc2f4c08b/cmd/cmd.go#L860))

The format looks like this:

```json
{
  "model_info": {
    "general.architecture": "llama",
    "llama.context_length": 132000
  }
}
```

Once this PR is merged we could technically remove the old code
https://github.com/zed-industries/zed/blob/8c3f09e31e3588a2494042dfe77bb3dccab1f7ba/crates/ollama/src/ollama.rs#L22
I decided to keep it for now, as it is unclear if the necessary fields
are available via the API on older Ollama versions.

Release Notes:

- Fixed an issue where Ollama models would use the wrong context window
size

Closes #39863

Regressed in https://github.com/zed-industries/zed/pull/36653

Release Notes:

- Fixed an issue where clicking a sticky item in the project panel
wouldn’t correctly scroll the view to show its start.

Release Notes:

- N/A

Signed-off-by: Xiaobo Liu <cppcoffee@gmail.com>

Closes #39585 

Release Notes:

- N/A

The beta template will be removed after Windows launch, the new url will
be:


https://github.com/zed-industries/zed/issues/new?template=07_bug_windows.yml

Release Notes:

- N/A

This is a small step toward a future where all tree view item-like
elements in Zed can actually use this component.

Release Notes:

- N/A

The issue was caused by the scroll handle taking a couple of frames to
update its offset correctly after calling
`ScrollHandle::scroll_to_top_of_item`. The fast fix is forcing 3 frames
to render back-to-back.

In the future, we should look into `ScrollHandle` and see if there's any
way to update its state outside of paint.

Release Notes:

- N/A

Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
Co-authored-by: Mikayla Maki <mikayla.c.maki@gmail.com>
Co-authored-by: Katie Geer <katie@zed.dev>
Co-authored-by: Ben Kunkle <ben@zed.dev>

Now that the toggle nav focus works well, we can advertise it!

Release Notes:

- N/A

Closes #39875

Release Notes:

- Fixed "open/copy permalink to line" paths not being URL-escaped

Co-authored-by: Cole Miller <cole@zed.dev>

Closes #39883

Release Notes:

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

---------

Co-authored-by: Anthony <anthony@zed.dev>

Re-adds: https://github.com/zed-industries/zed/pull/37264

This PR re-adds basic support for showing HTML images, without touching
the display mode for images.
The initial PR changed the `div().flex().flex_col()` to
`h_flex().flex_wrap()` but this broke the text wrapping in almost all
cases.

**Note**: This does not add support for showing the images inline,
because we haven't figured out how they correctly do this.
I'm working on adding the CSS `inline` display feature support to taffy
that hopefully allows us to correctly show images/other elements inline
without breaking the text wrapping.

**Before (nightly) and after (dev) for the README file inside Zed.
(nothing has changed, which is good)**
<img width="3440" height="1380" alt="Screenshot 2025-09-13 at 12 49 08"
src="https://github.com/user-attachments/assets/9cbdcb07-dbe9-4236-9d20-e59acc0e955e"
/>

**Result**
<img width="1717" height="1314" alt="Screenshot 2025-09-13 at 12 51 54"
src="https://github.com/user-attachments/assets/1c0f8507-c63d-472e-8e82-a654a63f7153"
/>

cc @SomeoneToIgnore

Release Notes:

- markdown preview: Added support for HTML `img` tags inside paragraphs

As you can see in the image, we were previously returning different
`ascent`s/`descent`s when a line would/would not contain an Emoji.

<img width="104" height="36" alt="image"
src="https://github.com/user-attachments/assets/436aeda0-87c0-4dee-943b-6da83681d466"
/>

---
CoreTexts `CTLineGetTypographicBounds` seems to return a different
ascent/descent depending on if an Emoji is there or not AFAIK it is not
documented if this is intended behaviour or not. For us it is
undesirable, as typing an Emoji causes the line to be shifted to the
bottom, see here:


https://github.com/user-attachments/assets/2ad1c82e-6297-48ac-a522-fb382ea56eea

--- 
Instead of using `CTLineGetTypographicBounds` to resolve the
ascent/descent, we look at every run and choose the maximum
ascent/descent. This matches how it [works on
Linux](https://github.com/zed-industries/zed/blob/f1d17fcfbef41690fdeb523f0fbddc7c406c5ad6/crates/gpui/src/platform/linux/text_system.rs#L452)

Release Notes:

- Fixed an issue on macOS where typing an emoji on a line would cause
the line to shift downwards by a few pixels

Release Notes:

- N/A

Closes #38189 

- Fixed border dashed for diverse scenarios, as demonstrated in the
images below.
- This change has no impact on the rendering of solid borders, as it was
implemented inside an if block for dashed styles

Release Notes:
  - N/A

## Before Images
<details><summary>click to expand (small top border, medium right
border, large bottom border)</summary>
<img width="289" height="95" alt="Screenshot From 2025-09-15 13-28-14"
src="https://github.com/user-attachments/assets/5226cd0a-49c2-43b8-9df9-f64390e3759e"
/>
</details>
<details><summary> click to expand (Same size pairs of borders)
</summary>
<img width="289" height="95" alt="Screenshot From 2025-09-15 13-32-22"
src="https://github.com/user-attachments/assets/603e7b49-e8b1-45a4-ac35-1b3aedf52bca"
/>
<img width="289" height="95" alt="Screenshot From 2025-09-15 13-33-24"
src="https://github.com/user-attachments/assets/4243786c-4c9d-4419-91d6-4594b5ee4390"
/>
</details>

## After Images

<details><summary>click to expand (small top border, medium right
border, large bottom border)</summary>

<img width="289" height="95" alt="Screenshot From 2025-09-15 13-17-28"
src="https://github.com/user-attachments/assets/e2652b38-1c24-432e-b7fd-c6f4d4c71de6"
/>

</details>


<details><summary> click to expand (same size pairs of
borders)</summary>
<img width="289" height="95" alt="Screenshot From 2025-09-15 13-37-59"
src="https://github.com/user-attachments/assets/05228431-4a91-4531-adcd-d70acd2c3b44"
/>

<img width="289" height="95" alt="Screenshot From 2025-09-15 13-36-34"
src="https://github.com/user-attachments/assets/6da946b8-1ccd-4ed1-9b38-539eba4edf42"
/>
</details>

Closes #ISSUE

Release Notes:

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

Fixes the `Open Diff` action for untracked files when the `sort_by_path`
setting is enabled. The `ProjectDiff` wasn't correctly moving the
multibuffer's cursor to the untracked file because, when that setting is
enabled, it's sort prefix is changed to the tracked files sort prefix, and that
wasn't accounted for in `move_to_entry`.

Before these changes, the `sort_prefix` field for `PathKey` was called `namespace`, it was renamed to be clearer what its purpose is.

Closes #39529 

Release Notes:

- Fixed 'Open Diff' action for untracked files when `sort_by_path` is
enabled

---------

Co-authored-by: David Kleingeld <davidsk@zed.dev>

Release Notes:

- N/A

---------

Co-authored-by: Cole Miller <cole@zed.dev>

- Prefer agent-specific logout handling to allow state reset 
- Treat any auth method as supported; remove provider-specific filter 
- Avoid prompting auth when issuing /logout and agent supports it

Release Notes:

- N/A

Fixes: https://github.com/longbridge/gpui-component/issues/994

1. When SVG is rotated, incorrect graphics are drawn.

For example: the original aspect ratio of the SVG is 1:1, if the bounds
used to render the SVG are 400x200 (aspect ratio 2:1),
[here](https://github.com/zed-industries/zed/blob/21f985a018f7cca9c0fb7f5b7a87555486ab9db5/crates/gpui/src/svg_renderer.rs#L91)
the width is used as the scaling factor, causing the rendered SVG to
only have half the height. This PR ensures the complete SVG image is
always rendered.

2. The clipping region has no transformation applied, I added a function
called `distance_from_clip_rect_transformed` in the shader.

3. Fixed `monochrome_sprite_fragment` in `shader.metal` not applying
clipping region.

### Before:


https://github.com/user-attachments/assets/8f93ac36-281e-4837-96cd-c308bfbf92d1

### After:


https://github.com/user-attachments/assets/f52b67a6-4cb9-4d6c-b759-bbb91b59c1cf

Release Notes:

- N/A

---------

Co-authored-by: Jason Lee <huacnlee@gmail.com>

These changes refactor the whitespace handling logic for Vim's change
surrounds command (`cs`), making its behavior closely match
[tpope/vim-surround](https://github.com/tpope/vim-surround), following
[this
discussion](https://github.com/zed-industries/zed/issues/38169#issuecomment-3304129461).

Zed's current implementation has two main differences when compared to
[tpope/vim-surround](https://github.com/tpope/vim-surround):

- It only considers whether a single space should be added or removed,
instead of all the space that is between the surrounding character and
the content
- It only takes into consideration the new surrounding characters in
order to determine whether to add or remove that space

A review of
[tpope/vim-surround](https://github.com/tpope/vim-surround)'s behavior
reveals these rules for whitespace:

* Quote to Quote
    * Whitespace is never changed
* Quote to Bracket
    * If opening bracket, add one space
    * If closing bracket, do not add space
* Bracket to Bracket
    * If opening to opening, keep only one space
    * If opening to closing, remove all space
    * If closing to opening, add one space
    * If closing to closing, do not change space
* Bracket to Quote
    * If opening, remove all space
    * If closing, preserve all space

Below is a table with examples for each scenario. A new test has also
been added to specifically check the scenarios outlined above,
`vim::surrounds::test::test_change_surrounds_vim`.

| Type              | Before      | Command | After         |
|-------------------|-------------|---------|---------------|
| Quote → Quote     | `'   a   '` | `cs'"`  | `"   a   "`   |
| Quote → Quote     | `"   a   "` | `cs"'`  | `'   a   '`   |
| Quote → Bracket   | `'   a   '` | `cs'{`  | `{    a    }` |
| Quote → Bracket   | `'   a   '` | `cs'}`  | `{   a   }`   |
| Bracket → Bracket | `[   a   ]` | `cs[{`  | `{ a }`       |
| Bracket → Bracket | `[   a   ]` | `cs[}`  | `{a}`         |
| Bracket → Bracket | `[   a   ]` | `cs]{`  | `{    a    }` |
| Bracket → Bracket | `[   a   ]` | `cs]}`  | `{   a   }`   |
| Bracket → Quote   | `[   a   ]` | `cs['`  | `'a'`         |
| Bracket → Quote   | `[   a   ]` | `cs]'`  | `'   a   '`   |

These changes diverge from
[tpope/vim-surround](https://github.com/tpope/vim-surround) when
handling newlines. For example, with the following snippet:

```rust
fn test_surround() {
    if 2 > 1 {
        println!("place cursor here");
    }
};
```

Placing the cursor inside the string and running any combination of
‎`cs{[`, ‎`cs{]`, ‎`cs}[`, or ‎`cs}]` would previously remove newline
characters. With these changes, using commands like ‎`cs}]` will now
preserve newlines.

Related to #38169
Closes #39334

Release Notes:

- Improved Vim’s change surround command to closely match
[tpope/vim-surround](https://github.com/tpope/vim-surround) behavior.

---------

Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>

This takes the idea that @RemcoSmitsDev started on in
https://github.com/zed-industries/zed/pull/39354. We did away with
grabbing a snapshot of the display map when buffer coordinates were
sufficient.
Closes #37267

Release Notes:

- Reduced micro-stutters in project search with large multi-buffer
contents.

---------

Co-authored-by: Smit Barmase <heysmitbarmase@gmail.com>

Closes #39506

Release Notes:

- Fixed environment variable capture when login shell is
[xonsh](https://xon.sh/)

---------

Co-authored-by: Jakub Konka <kubkon@jakubkonka.com>

Release Notes:

- N/A

Release Notes:

- N/A

Release Notes:

- N/A

Closes #ISSUE

From notes:

```markdown
  - [x] Clicking on the disclsoure icon button in the root-level tree view item should steal focus and move it to the root item (not the icon button)
  - [x] [@ben] Allow left/right arrow keys to expand/collapse root tree view items in the nav
    - [x] With this, make enter/space work the same as clicking (activate page, don't expand root items, focus moves to the content and leaves nav — becomes consistent with mouse interaction)
  - [x] Smart cmd-shift-e: toggling focus should take you to the selected item
  - [x] [@ben] pageup + pagedown in nav -> jump between root items
  - [x] [@ben] home + end buttons should work
    - in nav:
      - home always goes to first section header
      - end always goes to last _visible_ item (does not expand)
```

Release Notes:

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