Gist is we only need to block the foreground thread for reparsing if
immediate language changes are useful to the user. That is usually only
the case when they edit the buffer
Release Notes:
- Improved performance of large project searches and project diffs
Co-authored by: David Kleingeld <david@zed.dev>
Summary
Fixed the thread deletion issue by:
1. Click handler conflict between the trash icon and the main ListItem
2. Added `cx.stop_propagation()` to prevent the click event from
bubbling up to the parent ListItem's click handler
3. Followed the established `cx.stop_propagation()` pattern used
throughout the codebase
Now when you click the trash icon to delete a thread:
- ✅ The thread deletion happens immediately on the first click
- ✅ No race condition between deletion and activation
- ✅ No double-clicking required
The fix is minimal, follows established patterns, and addresses the
exact root cause of the issue.
- Stop event propagation in thread removal handler
Release Notes:
- agent: Improved delete thread action in the history view by preventing
it from also triggering the thread activation.
Closes https://github.com/zed-industries/zed/issues/43721
This PR makes the loading state clearer by disabling the message editor
while the agent is loading as well as pulsating the icon.
Release Notes:
- agent: Made the thread loading state clearer in the agent panel.
Closes https://github.com/zed-industries/zed/issues/43728
Release Notes:
- agent: Fixed a bug where the plan and edit files list would consume
the whole space of the thread if they were too long. They're now capped
by a max-height and scrollable.
Sometimes we are unable to receive messages at all from an agent. This
puts on upper bound on the `initialize` call so we can at least give a
message to the user that something is wrong here.
30s might feel like too long, but I wanted to avoid some false positives
in case there was something an agent needed to do at startup. This will
still communicate to the user at some point that something is wrong,
rather than leave them waiting forever with no signal that something is
going wrong.
Release Notes:
- agent: Show an error message to the user if we are unable to
initialize an ACP agent in a reasonable amount of time.
We were using `std::path::Path::strip_prefix` to determine which
repository an absolute path belongs to, which doesn't work when the
paths are Windows-style but the code is running on unix. Replace it with
a platform-agnostic implementation of `strip_prefix`.
Release Notes:
- Fixed git features not working when a Windows host collaborates with a
unix guest
It is easy for us to get the two fields out of sync causing weird
problems, there is no reason to have both here so.
Release Notes:
- N/A *or* Added/Fixed/Improved ...
Co-authored by: Antonio Scandurra <antonio@zed.dev>
Follow up to #39021.
<img width="576" height="141" alt="image"
src="https://github.com/user-attachments/assets/c89885a4-e664-4614-9bb0-86442dff34ee"
/>
- Add migration to remove `source` tag because `ContextServerSettings`
is now untagged
- Fix typos in context server modal
- PR seems to have removed the `test_action_namespaces` test, which I
brought back in this PR
Release Notes:
- Fixed an issue where the `source` property of MCP settings would show
up as unrecognised
Closes https://github.com/zed-industries/zed/issues/42753
Consider the following flow: you submit prompt A. Prompt A generates
some edits. You don't click on either "reject" or "keep"; they stay in a
pending state. You then submit prompt B, but before the agent outputs
any response, you click to edit prompt B, thus submitting a
regeneration.
Before this PR, the above flow would make the edits originated from
prompt A to be auto-rejected. This feels very incorrect and can surprise
users when they see that the edits that were pending got rejected. It
feels more correct to only auto-reject changes if you're regenerating
the prompt that directly generated those edits in the first place. Then,
it also feels more correct to assume that if there was a follow-up
prompt after some edits were made, those edits were passively
"accepted".
So, this is what this PR is doing. Consider the following flow to get a
picture of the behavior change:
- You submit prompt A.
- Prompt A generates some edits.
- You don't click on either "reject" or "keep"; they're pending.
- You then submit prompt B, but before the agents outputs anything, you
click to edit prompt B, submitting a regeneration.
- Now, edits from prompt A will be auto-kept.
Release Notes:
- agent: Improved the "reject"/"keep" behavior when regenerating older
prompts by auto-keeping pending edits that don't originate from the
prompt to-be-regenerated.
This PR makes the thread markdown editable. This refers to the "open
thread as markdown" feature, where you previously could only read. One
benefit of this move is that it makes a bit more obvious that you can
`cmd-s` to save the markdown, allowing you to store the content of a
given thread. You could already do this before, but due to it being
editable now, you see the tab with a dirty indicator, which communicates
that better.
Release Notes:
- agent: Made the thread markdown editable.
Fixes a panic that was introduced in #42633. Repro steps:
1. Open the inline assistant and mention a file in the prompt
2. Run the inline assistant
3. Remove the mention and insert a different one
4. 💥
This would happen because the mention set still had a reference to the
old editor, because we create a new one in `PromptEditor::unlink`.
Also removes the unused
`crates/agent_ui/src/context_picker/completion_provider.rs` file, which
was not removed by mistake in the previous PR.
Release Notes:
- N/A
This PR adds back the footer with the "Configure" button in the model
selector but only when the seeing it from the Zed agent (or inline
assistant/text threads). I had removed it a while back because seeing
the "Configure" button, which takes you to the agent panel settings
view, when clicking from an external agent didn't make much sense, given
there's nothing model-wise you can configure from Zed (at least yet) for
an external agent.
This also makes the button in the footer a bit nicer by making it full
screen and displaying a keybinding, so that you can easily do the whole
"trigger model selector → go to settings view" all with the keyboard.
<img width="400" height="870" alt="Screenshot 2025-11-21 at 10 38@2x"
src="https://github.com/user-attachments/assets/c14f2acf-b793-4bc1-ac53-8a8a53b219e6"
/>
Release Notes:
- N/A
Closes#39088
Release Notes:
- Fixed AI assistant edits being scrambled when file was modified while
it was open
--
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
This allows you to specify default_model and default_mode for ACP
extensions, e.g.
```
"auggie": {
"default_model": "gpt-5",
"default_mode": "default",
"type": "extension"
},
```
Release Notes:
- Added support for specifying settings for ACP extensions
(`default_mode`, `default_model`)
Editor is a choke point in our compilation graph while also being a very
common crate that is being edited. So reducing things that depend on it
will generally improve compilation times for us.
Release Notes:
- N/A *or* Added/Fixed/Improved ...
This PR builds on top of the `default_mode` feature where it was
possible to set an external agent mode as the default if you held a
modifier while clicking on the desired option. Now, if you want to have,
for example, Haiku as your default Claude Code model, you can do that.
This feature adds parity between external agents and Zed's built-in one,
which already supported this feature for a little while.
Note: This still doesn't work with external agents installed from
extensions. At the moment, this is limited to Claude Code, Codex, and
Gemini—the ones we include out of the box.
Release Notes:
- agent: Added the ability to set a model as the default for a given
built-in external agent (Claude Code, Codex CLI, or Gemini CLI).
This PR introduces a new `MultiBufferOffset` new type wrapping size. The
goal of this is to make it clear at the type level when we are
interacting with offsets of a multi buffer versus offsets of a language
/ text buffer. This improves readability of things quite a bit by making
it clear what kind of offsets one is working with while also reducing
accidental bugs by using the wrong kin of offset for the wrong API.
This PR also uncovered two minor bugs due to that.
Does not yet introduce the MultiBufferPoint equivalent, that is for a
follow up PR.
Release Notes:
- N/A *or* Added/Fixed/Improved ...
### What this solves
This PR adds support for HTTP and SSE (Server-Sent Events) transports to
Zed's context server implementation, enabling communication with remote
MCP servers. Currently, Zed only supports local MCP servers via stdio
transport. This limitation prevents users from:
- Connecting to cloud-hosted MCP servers
- Using MCP servers running in containers or on remote machines
- Leveraging MCP servers that are designed to work over HTTP/SSE
### Why it's important
The MCP (Model Context Protocol) specification includes HTTP/SSE as
standard transport options, and many MCP server implementations are
being built with these transports in mind. Without this support, Zed
users are limited to a subset of the MCP ecosystem. This is particularly
important for:
- Enterprise users who need to connect to centralized MCP services
- Developers working with MCP servers that require network isolation
- Users wanting to leverage cloud-based context providers (e.g.,
knowledge bases, API integrations)
### Implementation approach
The implementation follows Zed's existing architectural patterns:
- **Transports**: Added `HttpTransport` and `SseTransport` to the
`context_server` crate, built on top of the existing `http_client` crate
- **Async handling**: Uses `gpui::spawn` for network operations instead
of introducing a new Tokio runtime
- **Settings**: Extended `ContextServerSettings` enum with a `Remote`
variant to support URL-based configuration
- **UI**: Updated the agent configuration UI with an "Add Remote Server"
option and dedicated modal for remote server management
### Changes included
- [x] HTTP transport implementation with request/response handling
- [x] SSE transport for server-sent events streaming
- [x] `build_transport` function to construct appropriate transport
based on URL scheme
- [x] Settings system updates to support remote server configuration
- [x] UI updates for adding/editing remote servers
- [x] Unit tests using `FakeHttpClient` for both transports
- [x] Integration tests (WIP)
- [x] Documentation updates (WIP)
### Testing
- Unit tests for both `HttpTransport` and `SseTransport` using mocked
HTTP client
- Manual testing with example MCP servers over HTTP/SSE
- Settings validation and UI interaction testing
### Screenshots/Recordings
[TODO: Add screenshots of the new "Add Remote Server" UI and
configuration modal]
### Example configuration
Users can now configure remote MCP servers in their `settings.json`:
```json
{
"context_servers": {
"my-remote-server": {
"enabled": true,
"url": "http://localhost:3000/mcp"
}
}
}
```
### AI assistance disclosure
I used AI to help with:
- Understanding the MCP protocol specification and how HTTP/SSE
transports should work
- Reviewing Zed's existing patterns for async operations and suggesting
consistent approaches
- Generating boilerplate for test cases
- Debugging SSE streaming issues
All code has been manually reviewed, tested, and adapted to fit Zed's
architecture. The core logic, architectural decisions, and integration
with Zed's systems were done with human understanding of the codebase.
AI was primarily used as a reference tool and for getting unstuck on
specific technical issues.
Release notes:
* You can now configure MCP Servers that connect over HTTP in your
settings file. These are not yet available in the extensions API.
```
{
"context_servers": {
"my-remote-server": {
"enabled": true,
"url": "http://localhost:3000/mcp"
}
}
}
```
---------
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
This PR's goal is to improve discoverability of how Zed "remembers" the
currently selected agent when hitting `cmd-n` (or `ctrl-n`). Hitting
that binding starts a new thread with whatever agent is currently
selected.
In the example below, I am in a Claude Code thread and if I hit `cmd-n`,
a new, fresh CC thread will be started:
<img width="500" height="822" alt="Screenshot 2025-11-18 at 1 13@2x"
src="https://github.com/user-attachments/assets/d3acd1aa-459d-4078-9b62-bbac3b8c1600"
/>
Release Notes:
- agent: Improved discoverability of the `cmd-n` keybinding to create a
new thread with the currently selected agent.
Closes https://github.com/zed-industries/zed/issues/32701
Release Notes:
- agent: Fixed a bug where hitting the `NewThread` keybinding when
focused inside a terminal within the agent panel would create a new
terminal tab instead of a new thread.
This PR makes the description in the callout that display general errors
in the agent panel be rendered as markdown. This allow us to pass URLs
to these error strings that will be clickable, improving the overall
interaction with them. Here's an example:
<img width="500" height="396" alt="Screenshot 2025-11-14 at 11 43@2x"
src="https://github.com/user-attachments/assets/f4fc629a-6314-4da1-8c19-b60e1a09653b"
/>
Release Notes:
- agent: Improved the interaction with errors by allowing links to be
clickable.
Closes https://github.com/zed-industries/zed/issues/36330
Closes https://github.com/zed-industries/zed/issues/35460
This PR improves how we display markdown tables by relying on grids
rather than flexbox. Given this makes text inside each cell wrap, I
ended up removing the `table_overflow_x_scroll` method, as it was 1)
used only in the agent panel, and 2) arguably not the best approach as a
whole, because as soon as you need to scroll a table, you probably need
more elements to make it be really great.
One thing I'm slightly unsatisfied with, though, is the border
situation. I added a half pixel border to the cell so they all sum up to
1px, but there are cases where there's a tiny space between rows and I
don't quite know where that's coming from and how it happens. But I
think it's a reasonable improvement overall.
<img width="500" height="1248" alt="Screenshot 2025-11-13 at 7 05@2x"
src="https://github.com/user-attachments/assets/182b2235-efeb-4a61-ada2-98262967355d"
/>
Release Notes:
- agent: Improved table rendering in the agent panel, ensuring cell text
wraps, not going off-screen.
## Summary
Fixes#42342
When installing a dev extension with long installation instructions, the
configuration modal would overflow and users couldn't scroll to see the
full content or interact with buttons at the bottom.
## Solution
This PR adds a `ScrollHandle` to the `ConfigureContextServerModal` and
passes it to the `Modal` component, enabling the built-in modal
scrolling capability. This ensures all content remains accessible
regardless of length.
## Changes
- Added `ScrollHandle` import to the ui imports
- Added `scroll_handle: ScrollHandle` field to
`ConfigureContextServerModal` struct
- Initialize `scroll_handle` with `ScrollHandle::new()` when creating
the modal
- Pass the scroll handle to `Modal::new()` instead of `None`
## Testing
- Built the changes locally
- Tested with extensions that have long installation instructions
- Verified scrolling works and all content is accessible
- Confirmed no regression for extensions with short descriptions
Release Notes:
- Fixed scrolling issue in extension configuration modal when
installation instructions overflow the viewport
---------
Co-authored-by: Finn Evers <finn.evers@outlook.de>
<img width="395" height="444" alt="Screenshot 2025-11-11 at 4 04 57 PM"
src="https://github.com/user-attachments/assets/8da68721-6e33-4d01-810d-4aa1e2f3402d"
/>
Discussed with @danilo-leal and we're going with the "it's checked in
both places" design!
Closes#40910
Release Notes:
- Recommended AI models now still appear in their normal category in
addition to "Recommended:"
Lukas Wirth
Adam Huganir
Aero
Xipeng Jin
Danilo Leal
Scott Churchley
Bennet Bo Fenner
Ben Brandt
Jason Lee
Remco SmitsandDanilo Leal
Kirill BulatovandConrad Irwin
Richard Feldman
Piotr OsiewiczandConrad Irwin
Smit Barmase
Andrew FarkasandConrad Irwin
Artur ShirokovandConrad Irwin
R.AmoghandFinn Evers