Now when the agent reads images (which the tool now explicitly
advertises to agents that it is capable of; previously the tool said it
could only read text even though it can actually read images), we see
them in the thread, and also they are auto-expanded by default so you
can see them when scrolling through:
<img width="725" height="1019" alt="Screenshot 2026-01-06 at 2 57 11 PM"
src="https://github.com/user-attachments/assets/5c908bad-48f2-46c2-afaa-7f189a178e05"
/>
This also adds a visual regression test that verifies images render
correctly in the agent thread view.
Unlike our previous visual tests, this one only renders the agent panel,
not the entire Zed window.
The "screenshot" it generates (rendered to a Metal texture) is from
completely mocked/simulated data structures, and looks like this:
<img width="546" height="984" alt="Screenshot 2026-01-06 at 2 54 41 PM"
src="https://github.com/user-attachments/assets/89a0921f-59e9-4dfe-94b2-4c3b625a851b"
/>
## Changes
- **New visual test**: `agent_thread_with_image` renders an
`AcpThreadView` containing a tool call with image content (the Zed app
icon)
- **Test infrastructure**: Added `StubAgentServer` helper and required
feature flags for visual testing
- **Test-support API**: Added `expand_tool_call()` method to
`AcpThreadView` to allow expanding tool calls for visual testing
- **Baseline screenshot**: Included baseline image showing the Zed logo
rendered in a tool call output
## How to run
```bash
# Run the visual tests
cargo run -p zed --bin visual_test_runner --features visual-tests
# Update baselines if UI intentionally changed
UPDATE_BASELINE=1 cargo run -p zed --bin visual_test_runner --features visual-tests
```
Release Notes:
- N/A
---------
Co-authored-by: Zed Zippy <234243425+zed-zippy[bot]@users.noreply.github.com>
Co-authored-by: Amp <amp@ampcode.com>
Breadcrumb visual reorchestration which makes a couple of key changes:
- Breadcrumbs are now displayed in file headers for multibuffer views.
Singleton buffers are unchanged.
- Multibuffer views now have collapse/expand all buttons available in
the toolbar
- Search in multibuffer views now takes up less space, by occupying the
primary toolbar location erstwhile taken up by breadcrumbs
<img width="1721" height="823" alt="Screenshot 2025-12-22 at 4 15 41 PM"
src="https://github.com/user-attachments/assets/38e924f7-fe9e-4c83-84fb-2eee98137f43"
/>
<img width="1722" height="762" alt="Screenshot 2025-12-22 at 4 15 51 PM"
src="https://github.com/user-attachments/assets/ccc98ca3-75ce-4aca-bceb-890817b0f04d"
/>
Release Notes:
- Moved breadcrumbs to file headers in multibuffer views
- Added collapse/expand all options for multibuffer views
- Improved use of space for multibuffer search by taking the place of
the former breadcrumb toolbar location
---------
Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
Co-authored-by: Zed Zippy <234243425+zed-zippy[bot]@users.noreply.github.com>
### Closes
- Maybe https://github.com/zed-industries/zed/issues/38252 (there might
be something going on with NPX too)
- Also addresses
https://github.com/zed-industries/zed/pull/39021#issuecomment-3644818347
### Why
Some more involved MCP servers timeout often, especially on first setup.
I got tired of having to set timeouts manually per server. I also
noticed a timeout could not be set for http context servers, which
causes Context7 or GitHub's remote servers to timeout at 60s sometimes.
### Overview
MCP Timeout Configuration Feature
This PR adds additional configurable timeout settings for Model Context
Protocol servers including a global timeout and the addition of timeouts
for http servers, addressing issues where servers were timing out after
a fixed 60 seconds regardless of user needs.
**Key Features:**
- **Global timeout setting** (`context_server_timeout`) - default
timeout for all MCP servers (default: 60s, max: 10min)
- **Per-server timeout overrides** - individual servers can specify
custom timeouts via `timeout` field
- **Precedence hierarchy** - per-server timeout > global timeout >
default (60s)
- **Automatic bounds checking** - enforces 10-minute maximum to prevent
resource exhaustion
- **Support for both transports** - works with stdio and HTTP-based
context servers
- **Comprehensive test coverage** - 3 new tests validating global,
override, and stdio timeout behavior
- **Full backward compatibility** - existing configurations work
unchanged with sensible defaults
### Release Notes:
- Added the ability to configure timeouts for context server tool calls.
The new global `context_server_timeout` setting controls the default
timeout (default is 60s, max: 10min). Additionally, per-server timeouts
can be configured using the `timeout` field within servers defined in
the `"context_servers" setting
---------
Co-authored-by: Ben Kunkle <ben@zed.dev>
Add support for OpenAI's /responses endpoint for models that don't
support /chat/completions API. This enables compatibility with newer
model variants (`gpt-5-codex`, `gpt-5-pro`, `o3-pro`, etc) while
maintaining compatibility with existing configs
Changes:
- Add `supports_chat_completions` flag to model capabilities that
defaults to true for existing behavior
- Implement responses API client with streaming support as per [OpenAI
documentation](https://app.stainless.com/api/spec/documented/openai/openapi.documented.yml).
- Add `ResponseEventMapper` to convert responses events to completion
events for maintainer simplicity
- Update UI to allow toggling `chat_completions` capability
- Add `gpt-5-codex` model
Closes#38858
Release Notes:
- Added support for `gpt-5-codex` model
---------
Co-authored-by: Bennet Bo Fenner <bennet@zed.dev>
This PR reworks the (still feature-gated) side-by-side diff view to use
a different approach to representing the multibuffers on the left- and
right-hand sides.
Previously, these two multibuffers used identical sets of buffers and
excerpts, and were made to behave differently by adding a new knob to
the multibuffer controlling how diffs are displayed. Specifically, the
left-hand side multibuffer would filter out the added range of each hunk
from the excerpts using a new `FilteredInsertedHunk` diff transform, and
the right-hand side would simply not show the deleted sides of expanded
hunks. This approach has some problems:
- Line numbers, and actions that navigate by line number, behaved
incorrectly for the left-hand side.
- Syntax highlighting and other features that use the buffer syntax tree
also behaved incorrectly for the left-hand side.
In this PR, we've switched to using independent buffers to build the
left-hand side. These buffers are constructed using the base texts for
the corresponding diffs, and their lifecycle is managed by `BufferDiff`.
The red "deleted" regions on the left-hand side are represented by
`BufferContent` diff transforms, not `DeletedHunk` transforms. This
means each excerpt on the left represents a contiguous slice of a single
buffer, which fixes the above issues by construction.
The tradeoff with this new approach is that we now have to manually
synchronize excerpt ranges from the right side to the left, which we do
using `BufferDiffSnapshot::row_to_base_text_row`.
Release Notes:
- N/A
---------
Co-authored-by: cameron <cameron.studdstreet@gmail.com>
Co-authored-by: HactarCE <6060305+HactarCE@users.noreply.github.com>
Co-authored-by: Miguel Raz Guzmán Macedo <miguel@zed.dev>
Co-authored-by: Anthony <anthony@zed.dev>
Co-authored-by: Cameron <cameron@zed.dev>
This has been something we've wanted to do for a long time since docs
aside thus far have been hard-anchored either at the top or bottom of
the container that their trigger is sitting in. This PR introduces the
change so that they're centered with their trigger, regardless of
whether it's on a context menu or picker, by including a canvas element
in both the container and the trigger to calculate where the docs aside
should precisely sit on. Here's the result:
https://github.com/user-attachments/assets/8147ad05-1927-4353-991d-405631de67d0
Note that at the moment, docs aside are only visible through _hovering_,
and ideally, they should be available on both hover and selection
(keyboard nav). But I'll leave that for later.
Release Notes:
- N/A
There were several places adding a copy icon button, so thought of
encapsulating the logic to copy a given string into the clipboard (and
other small details like swapping the icon and tooltip if copied) into a
component, making it easier to introduce this sort of functionality in
the future, with fewer lines of code.
All it takes (for the simplest case) is:
```rs
CopyButton::new(your_message)
```
<img width="600" height="714" alt="Screenshot 2025-12-29 at 10 50@2x"
src="https://github.com/user-attachments/assets/e6949863-a056-4855-82d8-e4ffb5d62c90"
/>
Release Notes:
- N/A
This PR introduces the `MentionCrease` component, aimed at solving two
issues with mention creases in the agent panel:
- Previously, the mention crease was using a button with a regular size,
which is bigger than the default buffer font line height. That made the
crease look clipped and also overlapping with one another when in a
multiple line scenario where the creases would be on top of each other.
`MentionCrease` uses the window line height value to set the button
height, with a small one pixel vertical padding just for a bit of
spacing.
- Previously, given the crease used a `Label`, its font size wouldn't
scale if you changed the `agent_buffer_font_size` setting. Now,
`MentionCrease` uses that font size value, which makes regular text and
its text grow together as you'd expect.
Release Notes:
- agent: Fix a bug where mention creases didn't scale with
`agent_buffer_font_size` and got clipped/jumbled when rendered one above
the other.
Closes#23158
Release Notes:
- Added a right-click context menu for the thread view in the agent
panel.
---------
Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
This PR adds the ability to favorite models for external agents—writing
to the settings in the `agent_servers` key—as well as a handful of other
improvements:
- Make the cycling keybinding `alt-enter` work for the inline assistant
as well as previous user messages
- Better organized the keybinding files removing some outdated
agent-related keybinding definitions
- Renamed the inline assistant key context to "InlineAssistant" as
"PromptEditor" is old and confusing
- Made the keybindings to rate an inline assistant response visible in
the thumbs up/down button's tooltip
- Created a unified component for the model selector tooltip given we
had 3 different places creating the same element
- Make the "Cycle Favorited Models" row in the tooltip visible only if
there is more than one favorite models
Release Notes:
- agent: External agents also now support the favoriting model feature,
which comes with a handy keybinding to cycle through the favorite list.
This PR fixes two issues with regards to markdown codeblocks rendered in
tool call input and output content display:
- the JSON code snippets weren't properly indented
- codeblocks weren't being rendered in unique containers; e.g., if you
hovered one scrollbar, all of them would also be hovered, even though
horizontal scrolling itself worked properly
Here's the end result:
https://github.com/user-attachments/assets/3d6daf64-0f88-4a16-a5a0-94998c1ba7e2
Release Notes:
- agent: Fix scrollbar and JSON indentation for tool call input/output
content's markdown codeblocks.
The failure would happen if the current version of the file was open as
an editor. This happened because the git blob and current version of the
buffer would have the same `ProjectPath`.
The fix was adding a new `DiskState::Historic` variant to represent
buffers that are past versions of a file (usually a snapshot from
version control). Historic buffers don't return a `ProjectPath` because
the file isn't real, thus there isn't and shouldn't be a `ProjectPath`
to it. (At least with the current way we represent a project path)
I also change the display name to use the local OS's path style instead
of being hardcoded to Posix, and cleaned up some code too.
Release Notes:
- N/A
---------
Co-authored-by: Cole Miller <cole@zed.dev>
Co-authored-by: cameron <cameron.studdstreet@gmail.com>
Co-authored-by: xipengjin <jinxp18@gmail.com>
In the case of large vertical_scroll_margin, we could scroll up such
that the assistant was out of view. Now, keep it no lower than the
center of the editor.
Closes#18058
Release Notes:
- N/A
- Both the mode, profile, and model selectors have the option to cycle
through its options with a keybinding. In the tooltip that shows it, in
some of them the "Cycle Through..." label was at the top, and in others
at the bottom. Now it's all at the bottom.
- We used different language in different places for "going to a file".
The tool call edit card's header said "_Jump_ to File" while the edit
files list said "_Go_ to File". Now it's both "Go to File".
Release Notes:
- N/A
This PR simplifies how we display thread timestamps in the agent panel's
history view. For threads that are older-than-yesterday, we just show
how many days ago that thread was had in. Hovering over the thread item
shows you both the title and the full date, if needed (time and date).
<img width="450" height="786" alt="Screenshot 2025-12-18 at 5 24@2x"
src="https://github.com/user-attachments/assets/11416e9b-f1b0-4307-9db0-988a95a316a1"
/>
Release Notes:
- N/A
Follow up to a regression that happened when we introduced agent servers
that made everywhere displaying agent names use the extension name
instead of the display name. This has been since fixed in other places
and this PR now updates the agent panel's message editor, too:
| Before | After |
|--------|--------|
| <img width="1154" height="254" alt="Screenshot 2025-12-18 at 12 54
2@2x"
src="https://github.com/user-attachments/assets/5f3de9f9-4e11-42f6-90c2-56fc8cdff32e"
/> | <img width="1154" height="254" alt="Screenshot 2025-12-18 at 12
54@2x"
src="https://github.com/user-attachments/assets/46ed5c45-7e1d-4cc6-b219-b6cc19206d1b"
/> |
Release Notes:
- N/A
Follow up to #42982
Release Notes:
- agent: Allow pasting code without formatting via ctrl/cmd-shift-v.
- agent: Fixed an issue where pasting a single line of code would always
insert an @mention
Closes#44717
Sometimes, we show the user the agent's auth methods because we got an
AuthRequired error.
However, there are also several ways a user can choose to re-enter the
authentication flow even though they are still logged in.
This has caused some confusion with several users, where after logging
in, they type /login again to see if anything changed, and they saw an
"Authentication Required" warning.
So, I made a distinction in the UI if we go to this flow from a concrete
error, or if not, made the language less error-like to help avoid
confusion.
| Before | After |
|--------|--------|
| <img width="1154" height="446" alt="Screenshot 2025-12-18 at 10
54@2x"
src="https://github.com/user-attachments/assets/9df0d59a-2d45-4bfc-ba85-359dd1a4c8ae"
/> | <img width="1154" height="446" alt="Screenshot 2025-12-18 at 10
53@2x"
src="https://github.com/user-attachments/assets/73a9fb45-4e6f-4594-8795-aaade35b2a72"
/> |
Release Notes:
- N/A
---------
Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
Co-authored-by: Miguel Raz Guzmán Macedo <miguel@zed.dev>
Follow up to https://github.com/zed-industries/zed/pull/44297.
Initial implementation in ce884443f1c38ca8da9edf9fbbb8e7fd579452cb used
`Arc` to store the reference to the hash map inside the iterator while
keeping the lifetime static. The code was later simplified in
5151b22e2ea37fb457cf4f88b88fe4f315306074 to build the list eagerly but
the Arc was forgotten, although it became unnecessary.
cc @bennetbo
Release Notes:
- N/A
Also adjust the code for streaming tool use to always use a
rewrite_section; remove insert_here entirely.
Release Notes:
- N/A
Co-authored-by: Max Brunsfeld <maxbrunsfeld@gmail.com>
Follow-up of https://github.com/zed-industries/zed/pull/44887
Trims the worktree trust mechanism to the actual `worktree`s, so now
"global", workspace-level things like `prettier`, `NodeRuntime`,
`copilot` and global MCP servers are considered as "trusted" a priori.
In the future, a separate mechanism for those will be considered and
added.
Release Notes:
- N/A
Closes #ISSUE
Problem:
- The status bar’s pending keystroke indicator (shown next to --NORMAL--
in Vim mode) didn’t clear when focus moved to another context, e.g.
hitting g in the editor then clicking the Git panel. The keymap state
correctly canceled the prefix, but observers that render the indicator
never received a “pending input changed” notification, so the UI kept
showing stale prefixes until a new keystroke occurred.
Fix:
- The change introduces a `pending_input_changed_queued` flag and a new
helper `notify_pending_input_if_needed` which will flushes the queued
notification as soon as we have an App context. The
`pending_input_changed` now resets the flag after notifying subscribers.
Before:
https://github.com/user-attachments/assets/7bec4c34-acbf-42bd-b0d1-88df5ff099aa
After:
https://github.com/user-attachments/assets/2264dc93-3405-4d63-ad8f-50ada6733ae7
Release Notes:
- Fixed: pending keybinding prefixes on the status bar now clear
immediately when focus moves to another panel or UI context.
---------
Co-authored-by: Nathan Sobo <nathan@zed.dev>
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
Previously, the list of edit files had a double axis scroll issue
because the list itself scrolled vertically and each file row would
scroll horizontally, causing a bad UX. The horizontal scroll intention
was so that you could see the whole path, but I've included it in the
tooltip in case it becomes obscured due to a small panel width.
<img width="500" height="666" alt="Screenshot 2025-12-17 at 11 24@2x"
src="https://github.com/user-attachments/assets/ea87236d-f5c6-475a-bf66-1afae7a6ca05"
/>
Release Notes:
- agent: N/A
Just touching up the tooltip casing, colors, and a bit of spacing. Also
added the keybiniding to close the assistant. Maybe it was obvious
already but I don't think it hurts.
Release Notes:
- N/A
Similar to how you can use `shift-tab` to cycle through profiles/modes,
you can now use `alt-tab` to cycle through the language models you have
favorited.
<img width="500" height="312" alt="Screenshot 2025-12-16 at 5 23@2x"
src="https://github.com/user-attachments/assets/006d417d-5da1-48f9-82cc-ea06e28adb30"
/>
Release Notes:
- agent: Added the ability to cycle through favorited models using the
`alt-tab` keybinding.
This PR solves my main pain point with Zed agent: I have a long list of
available models from different providers, and I switch between a few of
them depending on the context and the project. In particular, I use the
same models from different providers depending on whether I'm working on
a personal project or at my day job. Since I only care about a few
models (none of which are in "recommended") that are scattered all over
the list, switching between them is bothersome, even using search.
This change adds a new option in `settings.json`
(`agent.favorite_models`) and the UI to manipulate it directly from the
list of available models. When any models are marked as favorites, they
appear in a dedicated section at the very top of the list. Each model
has a small icon button that appears on hover and allows to toggle
whether it's marked as favorite.
I implemented this on the UI level (i.e. there's no first-party
knowledge about favorite models in the agent itself; in theory it could
return favorite models as a group but it would make it harder to
implement bespoke UI for the favorite models section and it also
wouldn't work for text threads which don't use the ACP infrastructure).
The feature is only enabled for the native agent but disabled for
external agents because we can't easily map their model IDs to settings
and there could be weird collisions between them.
https://github.com/user-attachments/assets/cf23afe4-3883-45cb-9906-f55de3ea2a97
Closes https://github.com/zed-industries/zed/issues/31507
Release Notes:
- Added the ability to mark language models as favorites and pin them to
the top of the list. This feature is available in the native Zed agent
(including text threads and the inline assistant), but not in external
agents via ACP.
---------
Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
Co-authored-by: Bennet Bo Fenner <bennetbo@gmx.de>
Closes#32951
## Summary
When an agent notification was shown and the `AcpThreadView` was dropped
(e.g., by closing the project window or navigating to a new thread), the
notification would become orphaned and undismissable because the
subscriptions handling dismiss events were dropped along with the thread
view.
## Fix
Added an `on_release` callback that closes all notification windows when
the thread view is dropped. This ensures notifications are always
cleaned up properly.
## Testing
Added `test_notification_closed_when_thread_view_dropped` to verify
notifications are closed when the thread view is dropped.
Release Notes:
- Fixed agent notification getting stuck and becoming undismissable when
the project window is closed or when navigating to a new thread
Closes https://github.com/zed-industries/zed/issues/12589
Forces Zed to require user permissions before running any basic
potentially dangerous actions: parsing and synchronizing
`.zed/settings.json`, downloading and spawning any language and MCP
servers (includes `prettier` and `copilot` instances) and all
`NodeRuntime` interactions.
There are more we can add later, among the ideas: DAP downloads on
debugger start, Python virtual environment, etc.
By default, Zed starts in restricted mode and shows a `! Restricted
Mode` in the title bar, no aforementioned actions are executed.
Clicking it or calling `workspace::ToggleWorktreeSecurity` command will
bring a modal to trust worktrees or dismiss the modal:
<img width="1341" height="475" alt="1"
src="https://github.com/user-attachments/assets/4fabe63a-6494-42c7-b0ea-606abb1c0c20"
/>
Agent Panel shows a message too:
<img width="644" height="106" alt="2"
src="https://github.com/user-attachments/assets/0a4554bc-1f1e-455b-b97d-244d7d6a3259"
/>
This works on local, SSH and WSL remote projects, trusted worktrees are
persisted between Zed restarts.
There's a way to clear all persisted trust with
`workspace::ClearTrustedWorktrees`, this will restart Zed.
This mechanism can be turned off with settings:
```jsonc
"session": {
"trust_all_worktrees": true
}
```
in this mode, all worktrees will be trusted by default, allowing all
actions, but no auto trust will be persisted: hence, when the setting is
changed back, auto trusted worktrees will require another trust
confirmation.
This settings switch was added to the onboarding view also.
Release Notes:
- Introduced worktree trust mechanism, can be turned off with
`"session": { "trust_all_worktrees": true }`
---------
Co-authored-by: Matt Miller <mattrx@gmail.com>
Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
Co-authored-by: John D. Swanson <swanson.john.d@gmail.com>
Fixes#43165
## Problem
MCP prompts were only available in text threads, not agent threads.
Users with MCP servers that expose prompts couldn't use them in the main
agent panel.
## Solution
Added MCP prompt support to agent threads by:
- Creating `ContextServerPromptRegistry` to track MCP prompts from
context servers
- Subscribing to context server events to reload prompts when MCP
servers start/stop
- Converting MCP prompts to available commands that appear in the slash
command menu
- Integrating prompt expansion into the agent message flow
## Testing
Tested with a custom MCP server exposing `explain-code` and
`write-tests` prompts. Prompts now appear in the `/` slash command menu
in agent threads.
Release Notes:
- Added MCP prompt support to agent threads. Prompts from MCP servers
now appear in the slash command menu when typing `/` in agent threads.
---------
Co-authored-by: Agus Zubiaga <agus@zed.dev>
Closes#26823
Release Notes:
- Added support for customising the prompt used for generating commit
message in the rules library
---------
Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
This PR introduces a few components for the model selector pickers.
Given we're still maintaining two flavors of it due to one of them being
wired through ACP and the other through the language model registry,
having one source of truth for the UI should help with maintenance
moving forward, considering that despite the internal differences, they
look and behave the same from the standpoint of the UI.
Release Notes:
- N/A
Lukas Wirth