This mirrors VSCode setting that inspired `snippet_sort_order` to begin
with; VSCode supports inline/top/bottom/none, with none completely
disabling snippet completion. See
https://code.visualstudio.com/docs/editing/intellisense#_snippets-in-suggestions
This is helpful for LSPs that do not allow configuring snippets via
configuration such as clangd.
Release Notes:
- Added `none` as one of the values for `snippet_sort_order` to
completely disable snippet completion.
Closes#31805
This is an issue with Linux currently that `window.focus` is `None` upon
startup in both X11 and Wayland. Specifically, the order in which
[this](8d05a3d389/crates/gpui/src/window.rs (L3116))
and
[this](8d05a3d389/crates/gpui/src/app.rs (L956))
are executed varies between Linux and macOS. That is, one tries to
remove (blur) focus from a window, while other checks window focus to
put that focus id to a frame. In macOS, blur happens afterwards setting
focus on a frame, but in Linux, the inverse of it happens, leading to
`window.focus` to `None`.
For the time being, we handle all visible buttons to take care of this
**focus can be `None`** case, and make it work anyway. But, we should
look at the deeper issue mentioned above with GPUI. Created new issue to
track that https://github.com/zed-industries/zed/issues/34591.
Release Notes:
- Fixed an issue where button clicks wouldn’t work on startup until
clicked on the center pane on Linux.
Keystroke input now gets cleared when toggling to normal search mode
Main search bar is focused when toggling to normal search mode
This also gets rid of highlight on focus from keystroke_editor because
it also matched the search bool field and was redundant
Release Notes:
- N/A
@probably-neb I guess we should be opening the keymap editor from title
bar and menu as well. I believe this got missed in this: #34568.
Release Notes:
- Open Keymap editor from settings from menu and title bar.
Google Gemini Docs now recommend usage of `GEMINI_API_KEY` and the
legacy `GOOGLE_AI_API_KEY` variable is no longer supported in the modern
SDKs.
Zed will now accept either.
Release Notes:
- N/A
This PR improves Keystroke search by:
1. Allow searching by modifiers without additional keys.
2. Take match count into consideration when deciding if we should show
an action as a search match.
3. Take order into consideration as well.
Release Notes:
- N/A
---------
Co-authored-by: Ben Kunkle <ben@zed.dev>
Closes #ISSUE
This change applies both to the UI (we render `<null>` as muted text
instead of `zed::NoAction`) as well as how we update the keymap file
(the duplicated binding is bound to `null` instead of `"zed::NoAction"`)
Release Notes:
- N/A *or* Added/Fixed/Improved ...
Regressed in https://github.com/zed-industries/zed/pull/33928
This PR clears the selection drag state when the editor focus is out.
To reproduce:
1. Select some item in buffer that has a go to definition.
2. Cmd+Click mouse down on it, but don't let go.
3. Wait for 300ms+.
4. Now cursor changed to green + (valid state, this is for selection
drag-n-drop).
5. Now let go of your mouse down, we switched to a different file.
Cursor looks normal.
6. Come back to the previous buffer, see green + cursor style (BUG!).
Release Notes:
- Fixed the issue where the green (+) cursor style sometimes appears
when navigating to the definition and then back to the previous buffer.
Closes #ISSUE
Creates a new language in the languages crate for the DSL used in Zed
keybinding context. Previously, keybind context was highlighted as Rust
in the keymap UI due to the expression syntax of Rust matching that of
the context DSL, however, this had the side effect of highlighting upper
case contexts (e.g. `Editor`) however Rust types would be highlighted
based on the theme. By extracting only the necessary pieces of the Rust
language `highlights.scm`, `brackets.scm`, and `config.toml`, and
continuing to use the Rust grammar, we get a better result across
different themes
Release Notes:
- N/A *or* Added/Fixed/Improved ...
We know have the ability to filter matches in the keymap editor search
by exact keystroke matches. This allows user's to have the same behavior
as vscode when they toggle all actions with the same bindings
We also fixed a bug where conflicts weren't counted correctly when
saving a keymapping. This cause issues where warnings wouldn't appear
when they were supposed to.
Release Notes:
- N/A
---------
Co-authored-by: Ben Kunkle <ben@zed.dev>
Gradually remove details from a patch to keep it within the size limit.
This helps avoid using too much context when the user pastes large
files, generates files, or just makes many changes between agent
notifications.
Release Notes:
- N/A
Closes#34314
This PR resolves an issue where serde(untagged) caused Rust None values
to serialize as null, which OpenRouter's Mistral API (when tool_choice
is present) incorrectly interprets as a defined value, leading to a 400
error. By replacing serde(untagged) with serde(snake_case), None values
are now correctly omitted from the serialized JSON, fixing the problem.
P.S. A separate PR will address serde(untagged) usage for other
providers, as null is not expected for them either.
Release Notes:
- Fix ToolChoice getting serialized to null on OpenRouter
Closes #ISSUE
This PR attempts to add workarounds for `use_key_equivalents` in the
keymap UI. First of all it makes it so that `use_key_equivalents` is
ignored when searching for a binding to replace so that replacing a
keybind with `use_key_equivalents` set to true does not result in a new
binding. Second, it attempts to infer the value of `use_key_equivalents`
off of a base binding when adding a binding by adding an optional `from`
parameter to the `KeymapUpdateOperation::Add` variant. Neither
workaround will work when the `from` binding for an add or the `target`
binding for a replace are not in the user keymap.
cc: @Anthony-Eid
Release Notes:
- N/A *or* Added/Fixed/Improved ...
Closes#34319
In this pr: https://github.com/zed-industries/zed/pull/33462 there was
check added for early return for active_thread and message_editor as
those are not present in the TextThread and only available in the Thread
the token count was not getting triggered for TextThread this pr fixes
that regression by moving the logic specific to Thread inside of thread
view match.
<img width="3024" height="1886" alt="CleanShot 2025-07-15 at 23 50
18@2x"
src="https://github.com/user-attachments/assets/bd74ae8b-6c37-4cdd-ab95-d3c253b8a948"
/>
Release Notes:
- Fix token count not getting shown in the TextThread
This change improves user/agent collaborative editing.
When the user edits files that are used by the agent, the
`project_notification` tool now pushes *diffs* of the changes, not just
file names. This helps the agent to stay up to date without needing to
re-read files.
Release Notes:
- Improved user/agent collaborative editing: agent now receives diffs of
user edits
Closes#26468#16667
This PR fixes the spacebar not working with multiple keyboard layouts on
Linux X11. I have tested this with Czech, Russian, German, German Neo 2,
etc. It seems to work correctly.
`XkbStateNotify` events correctly update XKB state with complete
modifier info (depressed/latched/locked), but `KeyPress/KeyRelease`
events immediately overwrite that state using `update_mask()` with only
raw X11 modifier bits. This breaks xkb state as we reset `latched_mods`
and `locked_mods` to 0, as well as we might not correctly handle cases
where this new xkb state needs to change.
Previous logic is flawed because `KeyPress/KeyRelease` event only gives
you depressed modifiers (`event.state`) and not others, which we try to
fill in from `previous_xkb_state`. This patch was introduced to fix
capitalization issue with Neo 2
(https://github.com/zed-industries/zed/pull/14466) and later to fix
wrong keys with German layout
(https://github.com/zed-industries/zed/pull/31193), both of which I have
tested this PR with.
Now, instead of manually managing XKB state, we use the `update_key`
method, which internally handles modifier states and other cases we
might have missed.
From `update_key` docs:
> Update the keyboard state to reflect a given key being pressed or
released.
>
> This entry point is intended for programs which track the keyboard
state explictly (like an evdev client). If the state is serialized to
you by a master process (like a Wayland compositor) using functions like
`xkb_state_serialize_mods()`, you should use `xkb_state_update_mask()`
instead. **_The two functins should not generally be used together._**
>
> A series of calls to this function should be consistent; that is, a
call with `xkb::KEY_DOWN` for a key should be matched by an
`xkb::KEY_UP`; if a key is pressed twice, it should be released twice;
etc. Otherwise (e.g. due to missed input events), situations like "stuck
modifiers" may occur.
>
> This function is often used in conjunction with the function
`xkb_state_key_get_syms()` (or `xkb_state_key_get_one_sym()`), for
example, when handling a key event. In this case, you should prefer to
get the keysyms *before* updating the key, such that the keysyms
reported for the key event are not affected by the event itself. This is
the conventional behavior.
Release Notes:
- Fix the issue where the spacebar doesn’t work with multiple keyboard
layouts on Linux X11.
when `terminal.detect_venv.activate_script` setting is default, pick the
appropriate activate script as per the `terminal.shell` settings
specified by the user. Previously when the activate_script setting is
default, zed always try to use the `activate` script, which only works
when the user shell is `bash or zsh`. But what if the user is using
`fish` shell in zed?
Release Notes:
- python: value of `activate_script` setting is now automatically
inferred based on the kind of shell the user is running with.
---------
Co-authored-by: Piotr Osiewicz <24362066+osiewicz@users.noreply.github.com>
A nicer way to visualize in which plan you're in and a bit of
personalization by adding the GitHub handle you're signed with in the
user menu, as a complement to the avatar photo itself. Taking advantage
of the newly added Chip component.
<img width="320" height="476" alt="CleanShot 2025-07-16 at 1 33 08@2x"
src="https://github.com/user-attachments/assets/36718a42-27d1-499e-ac81-1eef2cd00347"
/>
Release Notes:
- N/A
Also removing the `icon` field as the banner component always renders
with an icon anyway. Hopefully, this fixes any weird text wrapping that
was happening before.
Release Notes:
- N/A
Previously we would append `js-debug/src/dapDebugServer.js` to the value
of the `dap.JavaScript.binary` setting and `src/debugpy/adapter` to the
value of the `dap.Debugpy.binary` setting, which isn't particularly
intuitive. This PR fixes that.
Release Notes:
- debugger: Made the semantics of the `dap.$ADAPTER.binary` setting more
intuitive for the `JavaScript` and `Debugpy` adapters. In the new
semantics, this should be the path to `dapDebugServer.js` for
`JavaScript` and the path to the `src/debugpy/adapter` directory for
`Debugpy`.
---------
Co-authored-by: Remco Smits <djsmits12@gmail.com>
Right now it doesn't work at all (the PID doesn't get set in the
generated scenario), and it's sort of redundant with the picker
functionality.
Release Notes:
- N/A
This PR adds action `editor: toggle focus` which focuses to last active
editor pane item in workspace.
Release Notes:
- Added `editor: toggle focus` action, which focuses to last active
editor pane item.
---------
Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
Closes#33945. Here's my attempt to describe what's going on in that
issue and what this fix is doing:
We always render the terminal inline assistant starting on the line
after the cursor, with a height of 4 lines. When deploying it, we scroll
the viewport to the bottom of the terminal so that the assistant will be
in view.
When scrolling while the assistant is deployed (including in that case),
we need to make an adjustment that "pushes up" the terminal content by
the height of the assistant, so that we can scroll to see all the normal
content plus the assistant itself. That quantity is `scroll_top`, which
represents _how much height in the current viewport is occupied by the
assistant that would otherwise be occupied by terminal content_. So when
you scroll up and a line of the assistant's height goes out of view,
`scroll_top` decreases by 1, etc.
When we scroll to the bottom after deploying the assistant, we set
`scroll_top` to the result of calling `max_scroll_top`, which computes
it this way:
```
block.height.saturating_sub(viewport_lines.saturating_sub(terminal_lines))
```
Which, being interpreted, is "the height of the assistant, minus any
viewport lines that are not occupied by terminal content", i.e. the
assistant is allowed to eat up vertical space below the last line of
terminal content without increasing `scroll_top`.
The problem comes when we clear the screen---this adds a full screen to
`terminal_lines`, but the cursor is positioned at the top of the
viewport with blank lines below, just like at the beginning of a session
when `terminal_lines == 1`. Those blank lines should be available to the
assistant, but the `scroll_top` calculation doesn't reflect that.
I've tried to fix this by basing the `max_scroll_top` calculation on the
position of the cursor instead of the raw `terminal_lines` value. There
was also a special case for `viewport_lines == terminal_lines` that I
think can now be removed.
Release Notes:
- Fixed the positioning of the terminal inline assistant when it's
deployed after clearing the terminal.
In #33275 I was very conservative about when to retry when there are
errors in language completions in the Agent panel.
Now we retry in more scenarios (e.g. HTTP 5xx and 4xx errors that aren't
in the specific list of ones that we handle differently, such as 429s),
and also we show a notification if the thread halts for any reason.
<img width="441" height="68" alt="Screenshot 2025-07-15 at 12 51 30 PM"
src="https://github.com/user-attachments/assets/433775d0-a8b2-403d-9427-1e296d164980"
/>
<img width="482" height="322" alt="Screenshot 2025-07-15 at 12 44 15 PM"
src="https://github.com/user-attachments/assets/5a508224-0fe0-4d34-9768-25d95873eab8"
/>
Release Notes:
- Automatic retry for more Agent errors
- Whenever the Agent stops, play a sound (if configured) and show a
notification (if configured) if the Zed window was in the background.
Adds keyboard navigation to the keybind edit modal. Using up/down arrows
to select the previous/next input editor, and `cmd-enter` to save +
`escape` to exit
Release Notes:
- N/A *or* Added/Fixed/Improved ...
Closes#34044
`EditorMode::SingleLine { auto_width: true }` was only used for the
title editor in the rules library, and following
https://github.com/zed-industries/zed/pull/31994 we can replace that
with a normal single-line editor without problems. The auto-width editor
was interacting badly with the recently-added newline visualization
code, causing a panic during layout---by switching it to
`Editor::single_line` the newline visualization works there too.
Release Notes:
- Fixed a panic that could occur when opening the rules library.
---------
Co-authored-by: Finn <finn@zed.dev>
