Doesn't support editing or deleting comments yet, but inputs work:
<img width="797" height="881" alt="Screenshot 2026-01-14 at 11 51 36 AM"
src="https://github.com/user-attachments/assets/746e085b-a919-475a-b804-ee0377e40a0b"
/>
<img width="698" height="178" alt="Screenshot 2026-01-14 at 11 47 51 AM"
src="https://github.com/user-attachments/assets/3edac0e8-62e5-4af5-a324-df2bd293f3ca"
/>
<img width="695" height="695" alt="Screenshot 2026-01-14 at 11 47 58 AM"
src="https://github.com/user-attachments/assets/64795ffd-62b4-48ea-a46c-7724221095d7"
/>
### Features implemented:
- **Local comment storage**: Comments are stored per-hunk in the Editor,
with chronological ordering
- **Expandable comments section**: Shows "N Comments" header that can
expand/collapse
- **Inline editing**: Click edit to transform a comment row into an
editable text field with confirm/cancel buttons
- **Delete functionality**: Remove individual comments
- **Send Review to Agent**: Toolbar button with badge showing comment
count, batch-submits all comments across all files/hunks to the Agent
panel
- **User avatars**: Shows the current user's avatar (from their Zed
account) next to comments, falls back to Person icon when not logged in
- **Visual tests**: Added tests for one comment, multiple comments
expanded, and comments collapsed
Note that this is behind a feature flag, so no release notes yet.
Release Notes:
- N/A
---------
Co-authored-by: Zed Zippy <234243425+zed-zippy[bot]@users.noreply.github.com>
Co-authored-by: David Baldwin <baldwindavid@gmail.com>
Release Notes:
- Agent: Added a menu item in the agent panel right-click menu for
copying a given agent's response.
---------
Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
## Problem
Profiling showed that during agent streaming, `thread.rs:1393:23` was
appearing constantly as a hotspot. The issue was that every single token
from the model triggered:
1. `this.update(cx, ...)`
2. `handle_text_event()` (or thinking/redacted_thinking)
3. `cx.notify()`
4. UI re-render
This created significant foreground thread pressure, contributing to
~500ms delays visible in the profiler.
## Solution
Batch all immediately-available events using `now_or_never()` and
process them in a single `update()` call with one `notify()` at the end.
This approach is deterministic - it processes exactly what's available
right now, adapting naturally to network speed:
- When tokens arrive slowly, you get one at a time
- When they arrive in bursts, they batch automatically
## Changes
- Remove `cx.notify()` from `handle_text_event`,
`handle_thinking_event`, and `handle_redacted_thinking_event`
- Batch events in the streaming loop using `now_or_never()`
- Call `cx.notify()` once per batch instead of per-event
- Keep `cx.notify()` in `handle_tool_use_event` for immediate tool
feedback
Release Notes:
- Improve streaming tool call performance by batching UI updates.
Closes#45755
Rainbow brackets in HTML files do not really make sense - a tag is
always enclosed in angled brackets. With our current approach, that
means that we will highlight all brackets in HTML files with the same
color, because well, these angled brackets do not quite behave like in
other languages.
Hence, let themes color this again and just exluce HTML angled brackets
from rainbow bracket colorization, as it is no real rainbow bracket
colorization anyway.
Release Notes:
- N/A
Closes#40322
When users run multi-line terminal commands, the confirmation dialog was
only showing the first line with '- N more lines' truncated. This meant
dangerous commands like 'rm -rf' or 'sudo' operations could be hidden
from view, creating a security risk.
Now terminal commands always display in full, with scroll support for
very long commands (20+ lines) to keep the confirmation buttons visible.
Other tools continue using truncation for better UI efficiency since
they don't pose the same security concerns.
Release Notes:
- Agent: Fixed terminal command truncation in the agent panel to better
expose long commands (e.g., potentially dangerous multi-line commands
are now fully visible before execution)
---------
Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
Co-authored-by: dino <dinojoaocosta@gmail.com>
This PR makes queueing a prompt the default behavior when sending a new
one while there's an on-going generation. You can still send a prompt
that will immediately interrupt the agent with the `cmd-shift-enter`
keybinding, though, which preserves the current behavior.
The main motivation for this change is to make the queueing not only
more discoverable, but more useful as well, as we're parting from the
assumption that most of the time, what you want is to queue it as
opposed to interrupting it (even though it's still possible to do either
through the keybinding I mentioned above or simply by stopping the
generation and sending a new one).
Here's a quick video:
https://github.com/user-attachments/assets/37f92433-70ef-459f-98ff-41ed80e3e43f
In the video, I show sending one prompt and then sending two others that
fall straight into the queue. Then, in the middle of the generation of
my first prompt, I use the `cmd-shift-enter` keybinding to send a prompt
immediately, interrupting the agent, effectively being sent in front of
the queue.
Release Notes:
- Agent: Made queueing prompts the default behavior when sending them
while there's an on-going generation.
- **copilot: Fix double lease panic when signing out**
- **Extract copilot_chat into a separate crate**
- **Do not use re-exports from copilot**
- **Use new SignIn API**
- **Extract copilot_ui out of copilot**
Closes#7501
Release Notes:
- Fixed Copilot providing suggestions from different Zed windows.
- Copilot edit predictions now support jumping to unresolved
diagnostics.
Feature flagged for now as we test this out with actual agents to see if
we need to provide any more feedback to the RFD before committing to the
current setup.
Release Notes:
- N/A
This makes sure that all of the work we do for caching/refreshing
session history is reused everywhere we need to access the list.
Very important for external agents so we aren't rerequesting history on
every search or recent list update. Puts the reloading of history in one
place
Release Notes:
- N/A
Fixes subword motion incorrectly jumping to next line when near end of
line. Updates boundary detection to use exclusive boundaries with
need_next_char parameter, matching regular word motion behavior.
Refactors word and subword motion to share boundary detection logic.
Closes#17780
Release Notes:
- Fixed subword motion incorrectly jumping to the next line when near
the end of a line
---------
Co-authored-by: dino <dinojoaocosta@gmail.com>
This reverts commit 94faaebfec.
The logic changed in the [original
PR](https://github.com/zed-industries/zed/pull/45969) is either
misplaced or lacks a counterpart that reacts on `gpui::Image` drop one
way or another.
The change was dropping the texture out of the global rendering atlas
when an image preview tab was disposed of, while in reality, another
instance (agent panel or another image preview tab) could require the
very same atlas entry to be rendered meanwhile.
Currently, only `RetainAllImageCache` in Zed does any kind of image
cleanup, and any other image usages will leak memory.
What makes it harder is that the atlas entry needs to live as long as a
particular `Arc<Image>` lives, and some public `gpui` API expects this
type to stay:
e747cfc955/crates/gpui/src/platform.rs (L1851-L1866)
For image viewer in particular, we need to consider why
`RetainAllImageCache` is not used there; for the global, normal fix, we
need to consider ways to have `cx` and `window` and a way to react on
`Image` drop.
As an option, we could break the `gpui` API (as it [happens
periodically](https://github.com/zed-industries/zed/issues/46183)
already) and use `Entity<Image>` instead of `Arc`, then react with
`cx.on_release_in` for each such image.
Closes https://github.com/zed-industries/zed/issues/46755
Closes https://github.com/zed-industries/zed/issues/46435
Release Notes:
- Fixed panics on concurrent image handling
Release Notes:
- When copying and pasting from a multibuffer view (e.g. into the agent
panel), the line numbers from the selection corresponded to the line
numbers in the multibuffer view (incorrect), instead of the actual line
numbers from the file itself.
- I also handled the case in which the selection spans multiple excerpts
(different files in the multibuffer) by just returning None for that
case, but maybe it should instead be split into two selections?
Left is before (bugged, should be lines 8:16, instead it is 38:46),
right is after (fixed).
<div style="display:flex; gap:12px;">
<img
src="https://github.com/user-attachments/assets/b1bfce1d-8b6a-41c0-ac7e-51f7bd7b284e"
width="48%" />
<img
src="https://github.com/user-attachments/assets/2a4c33a0-a969-4a3e-9aa5-d2c2fefba3b2"
width="48%" />
</div>
Release Notes:
- Added two actions `move_to_start_of_larger_syntax_node` and
`move_to_end_of_larger_syntax_node` that move cursors to the start or
end of the parent tree-sitter node
Following up on my PR #41321, this PR only adds the actions that are
used to enable code navigation across syntax nodes, without binding them
to any keys (such as tab) by default. Both actions use the tree-sitter
syntax tree to find parent nodes of the nodes the cursors are currently
in. `move_to_start_of_larger_syntax_node` will then move each cursor to
the first position of the parent nodes while
`move_to_end_of_larger_syntax_node` to a position right after the parent
nodes.
Related issues and discussions: #22349, #14803, #42828, #13736.
This PR doesn't achieve "tab out" functionality in the exact sense as is
requested in these issues as it does not bind the actions to the tab
key. I hope this PR can start some discussion on what the best way
forward for these issues is. In the meantime, users can configure keys
to use these actions as they see fit to emulate "tab out" behavior. For
example,
```
"context": "Editor && vim_mode == insert && !in_snippet && !showing_completions",
"bindings": {
"tab": "editor::MoveToEndOfLargerSyntaxNode",
"shift-tab": "editor::Tab"
}
```
This will enable tab to skip past code structures like brackets when the
cursor is not in a snippet or the autocomplete menu is not open. At the
same time, shift tab will act as a backup tab.
## Summary
This PR fixes a issue where Chinese characters (and other IME-based
inputs) would duplicate in the editor on Linux Wayland after remote
desktop connected (e.g., via Sunshine) .
## Steps to Reproduce
requirement: linux with wayland, and Sunshine(this is remote desktop
server) installed
1. opening zed
2. using other device connect to Sunshine
3. input Chinese characters such as "的"
### Expected Result
"的"
### Actual Result
"的的"
### bug recording
https://github.com/user-attachments/assets/d18961ec-48cb-4c06-a396-9fc604591769
This one does `fim_prefix`, `fim_middle`, and `fim_suffix` in that
order, in the prompt, instead of putting the current middle last.
Release Notes:
- N/A
---------
Co-authored-by: Agus Zubiaga <agus@zed.dev>
Co-authored-by: Ben Kunkle <ben@zed.dev>
This PR introduces a project dropdown when working with multiple
folders/projects in one workspace. Here are some interaction details
that I hope improves the UX of working on this scenario significantly:
- The dropdown shows the currently "active" project, which is determined
by:
- Either the file you're currently editing
- Or the file you have just recently switched to
- Some example cases:
- If you are focused on file from project A but switch to project B in
the titlebar, nothing happens. However, as soon as you type on the file
from project A, the title bar will update and your active project will
return to being project A.
- If you're focused on file from project A and change tabs to a file
from project B, the title bar will update, showing project B as the
active one.
- The content you'll see in the branch picker will correspond to the
currently active project
- It's still possible to reach the "Recent Projects" picker through the
project dropdown
- It's possible to do all interactions (trigger dropdown, select active
project, and remove project from workspace) with the keyboard
https://github.com/user-attachments/assets/e2346757-74df-47c5-bf4d-6354623b6f47
Note that this entire UX is valid only for a multiple folder workspace
scenario; nothing changes for the single project case.
Release Notes:
- Workspace: Improved the UX of working with multiple projects in the
same workspace through introducing a project dropdown that more clearly
shows the currently active project as well as allowing you to change it.
This fixes an issue that we noticed in particular with Mercury edit
predictions.
* [x] fix storage to not go stale
* [x] exclude excerpts that intersect the cursor excerpt
* [x] see if string representation of excerpts can be cached, to avoid
rebuilding it on every prediction
Release Notes:
- N/A
---------
Co-authored-by: Ben Kunkle <ben@zed.dev>
This PR changes the keybinding for the `agent::RejectAll` and
`agent::KeepAll` actions. My initial goal was to free up `cmd-shift-n`
to open a new fresh window again, but that was being used for the reject
all action. Given I had to change the reject all, I figured I had to
change the keep all action, too, to keep them consistent.
Release Notes:
- Agent: Removed keybinding conflict (`cmd-shift-n`) between rejecting
all changes in the agent panel vs. opening a new fresh window.
This PR adds the ability to open local terminals when working in remote
projects. When working in a remote (often actually remoting into a local
container) I always need to run separate terminals outside Zed so I can
run local build tools, scripts, agents, etc. for the related project.
I'd like to be able to run all of these in the same Zed window and this
adds that ability via one-off local terminals.
## Changes
Adds an optional `local` parameter to terminal commands. When set to
`true`, creates a local shell on your machine instead of connecting to
the remote.
### Implementation
- Added `force_local` parameter to terminal creation logic
- Created `create_local_terminal()` method that bypasses remote client
- Updated terminal actions (`NewTerminal`, `NewCenterTerminal`) to
accept optional `local: bool` field (defaults to `false`)
### Usage
**Via keybinding:**
```json
{
"bindings": {
"cmd-t": "workspace::NewCenterTerminal",
"cmd-T": ["workspace::NewCenterTerminal", { "local": true }]
}
},
{
"context": "Terminal",
"bindings": {
"cmd-n": "workspace::NewTerminal",
"cmd-N": ["workspace::NewTerminal", { "local": true }],
},
},
```
**Behavior:**
- Default terminal commands continue to work as before (remote in remote
projects, local in local projects)
- The `local` parameter is optional and defaults to `false`
Release Notes:
- Added support for opening local terminals in remote projects via
`local` parameter on terminal commands.
## Problem:
When working with many `pinned tabs`, they consume significant
horizontal space in the tab bar, leaving less room for unpinned (active
working) tabs, especially on small screens. This creates a poor user
experience as users must constantly scroll to find their working tabs,
or the tabs become too narrow to read file names.
## Solution:
Added a new opt-in setting `tab_bar.show_pinned_tabs_in_separate_row`
that displays pinned and unpinned tabs in two separate rows:
- Top row: Pinned tabs with navigation buttons and tab bar controls
- Bottom row: Unpinned (working) tabs with full horizontal space
- when pinned and not pinned tabs we show 2 rows:
<img width="1512" height="159" alt="2_rows"
src="https://github.com/user-attachments/assets/62a85fcc-8073-40d2-a8e4-270efd0d0f21"
/>
This is a well-established UX pattern used in many popular IDEs
including JetBrains products (IntelliJ IDEA, WebStorm, PyCharm,
RubyMine), Visual Studio, and Eclipse.
The two-row layout only appears when both pinned AND unpinned tabs
exist. If only one type is present, a single row is displayed:
- when only not pinned tabs we show one row:
<img width="1510" height="111" alt="when only not pinned"
src="https://github.com/user-attachments/assets/048df4fc-4b17-4ea2-9c5b-b4db91cdfc78"
/>
- when only pinned tabs we show one row:
<img width="1510" height="111" alt="when only pinned tabs"
src="https://github.com/user-attachments/assets/7ccf2517-8711-4114-815d-6cbb1e89d7e2"
/>
**Important** - Default Behavior Preserved:
- The setting defaults to false - existing users will see no change
whatsoever
- Original tab bar logic is preserved in a dedicated
`render_single_row_tab_bar` method, while the new two-row layout lives
in `render_two_row_tab_bar`. Both methods share common components
(`configure_tab_bar_start`, `configure_tab_bar_end`,
`render_unpinned_tabs_container`, `render_tab_bar_drop_target`) ensuring
DRY code and consistent behavior.
- Purely opt-in - users must explicitly enable this feature via Settings
UI or settings.json
- No impact on default UX - Zed's default tab bar appearance and
behavior remains identical
This implementation ensures zero risk to the existing user experience
while providing an optional enhancement for users who prefer separated
tab rows.
Configuration
Via Settings UI:
- Open Settings (Cmd+,) → Editor section → "Pinned Tabs Layout" toggle
Via settings.json:
```json
{
"tab_bar": {
"show_pinned_tabs_in_separate_row": true
}
}
```
<img width="468" height="67" alt="manual_settings"
src="https://github.com/user-attachments/assets/d21f6dde-7683-47b3-8dca-d190049d32fb"
/>
## Video of implemented feature:
https://github.com/user-attachments/assets/b1e52074-a5a1-4c4e-ad59-e55e2213509d
## Tested:
- Verify default behavior unchanged (single row with pinned + unpinned
tabs inline)
- Enable setting → verify two rows appear when both pinned and unpinned
tabs exist
- Enable setting → verify single row when only pinned tabs exist
- Enable setting → verify single row when only unpinned tabs exist
- Verify drag & drop works in both layouts
- Verify navigation buttons and tab bar buttons render correctly
- Verify setting toggle works in Settings UI
- Verify setting works via settings.json
Tested on: MacBook Pro M4, macOS 26.2
## Release Notes:
- Added `tab_bar.show_pinned_tabs_in_separate_row` setting to display
pinned and unpinned tabs in separate rows, giving unpinned tabs full
horizontal space
---------
Co-authored-by: Matt Miller <mattrx@gmail.com>
Fix a bug with `:norm` that disallowed trailing whitespace.
Commands that accept generic args (defined with `args()`) now preserve
trailing whitespace, while commands that accept filenames (defined with
`filename()`) have whitespace pre-trimmed. This allows, for example,
`:norm I ` to correctly insert spaces, matching NeoVim's behavior.
Release Notes:
- vim: Fixed `:norm` command to preserve trailing whitespace in
arguments (e.g., `:norm I ` now correctly inserts two spaces)
---------
Co-authored-by: dino <dinojoaocosta@gmail.com>
Collab projects are considered trusted and remote clients use the
editor-related settings only. A better fix would be sync "not trusted"
indicator with the project host, and disallow clients' iteraction with
that indicator.
Release Notes:
- Fixed collab settings sync causing "not trusted" pop ups for client
When using the `editor::actions::CopyAndTrim` action with a multi-line
selection in vim's Visual Line mode, pasting would crash Zed.
The bug occurred because trimming splits a selection into per-line
ranges, creating multiple `editor::ClipboardSelection` entries. However,
when `is_entire_line` was true (Visual Line mode), no newline separators
were added between these entries in the clipboard text. The paste code
then assumed separators existed and read past the end of the text.
The fix ensures newline separators are always added between trimmed line
ranges, regardless of whether the original selection was in line mode.
Closes#46616
Release Notes:
- Fixed a crash when pasting after using `editor: copy and trim` in
vim's Visual Line mode
