The crash was caused by not accounting for the fact that a range of
collapse frames only counts as one entry. Causing the filter indices to
overshoot for indices after collapse frames (it was counting all
collapse frames instead of just one).
The test missed this because it all happened in one `cx.update` closure
and didn't render the stack frame list when the filter was applied. The
test has been updated to account for this.
Release Notes:
- N/A
Co-authored-by: Cole Miller <cole@zed.dev>
Since Claude Code has it's own preferred method of grabbing API keys, we
don't want to reuse this one.
Release Notes:
- acp: Don't share Anthropic API key from the Anthropic provider to
allow default Claude Code login options
---------
Co-authored-by: Agus Zubiaga <agus@zed.dev>
This reverts commit bf5ed6d1c9.
We believe this may be breaking some users whose shell initialization
scripts change the working directory.
Release Notes:
- N/A
Related to #37213, #37150
When listing previously-downloaded versions of an external agent, don't
try to use any downloads that are missing the agent entrypoint
(indicating that they're corrupt/unusable), and delete those versions,
so that we can attempt to download the latest version again.
Also report clearer errors when failing to start a session due to an
agent server entrypoint or root directory not existing.
Release Notes:
- N/A
We don’t know the background color behind a non-opaque editor, so we
should skip contrast correction in that case. This prevents
single-editor mode (which is always transparent) from showing weird text
colors when text is selected.
We can’t account for the actual background during contrast correction
because we compute contrast outside gpui, while the actual color
blending happens inside gpui during drawing.
<img width="522" height="145" alt="image"
src="https://github.com/user-attachments/assets/6ee71475-f666-482d-87e6-15cf4c4fceef"
/>
Release Notes:
- Fixed an issue where Command Palette text looked faded when selected.
### First issue
In the scenario where you have an API key configured in Zed and you run
`/logout`, clicking on `Use Anthropic API Key` would show `Method not
implemented`.
This happened because we were only intercepting the `Use Anthropic API
Key` click if the provider was NOT authenticated, which would not be the
case when the user has an API key set.
### Second issue
When clicking on `Reset API Key` the modal would be dismissed even
though you picked no Authentication Method (which means you still would
be unauthenticated)
---
This PR fixes both of these issues
Release Notes:
- N/A
When the first visible worktree is a single-file worktree, we would
previously try to use the absolute path of that file as the root
directory for external agents, causing an error. This PR changes how we
handle this situation: we'll use the root of the first non-single-file
visible worktree if there are any, and if there are none, the parent
directory of the first single-file visible worktree.
Related to #37213
Release Notes:
- acp: Fixed being unable to run external agents when a single file (not
part of a project) was opened in Zed.
When doing stable/preview releases simultaneously there are two tags and
two branches pushed. Previously nix was attempting 1 job for each. Our
current mac parallelism is 4.
Can't easily test this. 🤷
Release Notes:
- N/A
We'll just SSH into the Windows runners and look for crashes there.
Reverts #35926
Release Notes:
- N/A
---------
Co-authored-by: Peter Tripp <petertripp@gmail.com>
Closes: #32668
- Add
[tool_call_timeout_millis](https://github.com/cline/cline/pull/1904)
field to ContextServerCommand, like in Cline
- Update ModelContextServerBinary to include timeout configuration
- Modify Client to store and use configurable request timeout
- Replace hardcoded REQUEST_TIMEOUT with self.request_timeout
- Rename REQUEST_TIMEOUT to DEFAULT_REQUEST_TIMEOUT for clarity
- Maintain backward compatibility with 60-second default
Release Notes:
- context_server: Add support for configurable timeout for MCP tool
calls
---------
Co-authored-by: Ben Brandt <benjamin.j.brandt@gmail.com>
- I believe this is caused by metal not being found due to it being on
the XcodeBeta path, not sure if there's a better fix for this but it'll
work until 26 is the latest release
Release Notes:
- N/A
Hi! As part of https://github.com/zed-extensions/ruby/issues/162 we
would like to rename HTML/ERB to HTML+ERB since it is more syntactically
correct to treat such language as ERB on top of HTML rather than HTML or
ERB.
To keep the user experience intact, we outlined the prerequisites in the
linked issue. This is the first PR that adds the HTML+ERB language name
to the list of enabled languages for the Emmet extension. We will do the
same for the Tailwind configuration in the Zed codebase. Once the new
versions of Emmet and Zed are released, we will merge the pull request
in the Ruby extension repository and release the updated version. After
that, we will remove the old HTML/ERB and YAML/ERB languages. Let me
know if that sounds good. Thanks!
Release Notes:
- N/A
Co-authored-by: Marshall Bowers <git@maxdeviant.com>
Closes#32758
Release Notes:
- Resolved an issue with the Ollama provider that caused requests to
fail with a 400 error for models that don't support tools. The tools
object is now only sent to compatible models to ensure successful
requests.
Follow up: #34532Closes#35434
Mostly fixes a issue were when the tool_choice is none it was getting
serialised as null. This was fixed for openrouter just wanted to follow
up and cleanup for other providers which might have this issue as this
is against the spec.
Release Notes:
- N/A
Follow up to https://github.com/zed-industries/zed/pull/30598
This PR introduces the `display_options` field in the
`CompletionResponse`, allowing a code context menu width to be
dynamically dictated based on its larger item. This will allow us to
have the @-mentions and slash commands completion menus in the agent
panel not be bigger than it needs to be. It may also be relevant/useful
in the future for other use cases.
For now, we set all instances of code context menus to use a fixed
width, as defined in the PR linked above, which means this PR shouldn't
cause any visual change.
Release Notes:
- N/A
Co-authored-by: Michael Sloan <mgsloan+github@gmail.com>
This PR removes the `http_client` feature from the `gpui` crate, as it
wasn't really doing anything.
It only controlled whether we depend on the `http_client` crate, but
from what I can tell we always depended on it anyways.
Obviates https://github.com/zed-industries/zed/pull/36615.
Release Notes:
- N/A
Improves the error handling for openrouter and adds automatic retry like
anthropic for few of the status codes.
Release Notes:
- Improves error messages for Openrouter provider
- Automatic retry when rate limited or Server error from Openrouter
This commit fixes an issue with how the `AnyBrackets` object was handled
with change surrounds (`cs`). With the keymap below, if one was to use
`csb{` with the text `(bracketed)` and the cursor inside the
parentheses, the text would not change.
```json
{
"context": "vim_operator == a || vim_operator == i || vim_operator == cs",
"bindings": {
"b": "vim::AnyBrackets"
}
}
```
Unfortunately there was no implementation for finding a corresponding
`BracketPair` for the `AnyBrackets` object, meaning that, when using
`cs` (change surrounds) the code would simply do nothing.
This commit updates this logic so as to try and find the nearest
surrounding bracket (parentheses, curly brackets, square brackets or
angle brackets), ensuring that `cs` also works with `AnyBrackets`.
Closes#24439
Release Notes:
- Fixed handling of `AnyBrackets` in vim's change surrounds (`cs`)
Before we were setting up lots of test setup regardless of if we were
actually going to be making real requests or not.
This will hopefully help with intermittent test errors we're seeing on
Windows in CI.
Release Notes:
- N/A
Closes#37249
We no longer show edit prediction when composing IME since it isn't
useful for unfinished alphabet.
Release Notes:
- Fixed edit predictions showing up during partial IME composition.
Closes#37211
Regressed in https://github.com/zed-industries/zed/pull/33305
Every time the terminal updates, we emit
`SearchEvent::MatchesInvalidated` to trigger a re-run of the buffer
search, which calls `clear_matches` to drop stale results.
https://github.com/zed-industries/zed/pull/33305 PR also cleared the
selection when clearing matches, which caused this issue. We could fix
it by only clearing matches and selection when they’re non-empty, but
it’s better to not clear the selection at all. This matches how the
editor behaves and keeps it consistent. This PR reverts that part of
code.
Release Notes:
- Fixed an issue where text selection was lost during continuous
terminal output.
If we get back slash-commands that aren't supported, tell the user that
this is the problem.
Release Notes:
- Improve error messages for unsupported ACP slash-commands
---------
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
Closes#30188Closes#34911Closes#34906
Many OpenAI-compatible providers do not automatically filter the tool
schema to comply with the underlying model's requirements; they simply
proxy the request. This creates issues, as models like **Gemini**,
**Grok**, and **Claude** (when accessed via LiteLLM on Bedrock) are
incompatible with Zed's default tool schema.
This PR addresses this by defaulting to a more compatible schema subset
instead of the full schema.
### Why this approach?
* **Avoids Poor User Experience:** One alternative was to add an option
for users to manually set the JSON schema for models that return a `400
Bad Request` due to an invalid tool schema. This was discarded as it
provides a poor user experience.
* **Simplifies Complex Logic:** Another option was to filter the schema
based on the model ID. However, as demonstrated in the attached issues,
this is unreliable. For instance, `claude-4-sonnet` fails when proxied
through LiteLLM on Bedrock. Reliably determining behavior would require
a non-trivial implementation to manage provider-and-model combinations.
* **Better Default Behavior:** The current approach ensures that tool
usage works out-of-the-box for the majority of cases by default,
providing the most robust and user-friendly solution.
Release Notes:
- Improved tool compatibility with OpenAI API-compatible providers
Signed-off-by: Umesh Yadav <git@umesh.dev>
Co-authored-by: Peter Tripp <peter@zed.dev>
Right now if you open Zed, and we deserialize an agent that's behind a
feature flag (e.g. CC), we don't restore it because the feature flag
check hasn't happened yet at the time we're deserializing (due to auth
not having finished yet).
This is a simple fix: assume that if you had serialized it in the first
place, you must have had the feature flag enabled, so go ahead and
reopen it for you.
Release Notes:
- N/A
This commit fixes a bug with Zed's vim mode surrounds plugin when
dealing with replacing pairs with quote and the contents between the
pairs had some whitespace within them.
For example, with the following string:
```
' str '
```
If one was to use the `cs'"` command, to replace single quotes with
double quotes, the result would actually be:
```
"str"
```
As the whitespace before and after the closing character was removed.
This happens because of the way the plugin decides whether to add or
remove whitespace after and before the opening and closing characters,
repsectively. For example, using `cs{[` yields a different result from
using `cs{]`, the former adds a space while the latter does not.
However, since for quotes the opening and closing character is exactly
the same, this behavior is not possible, so this commit updates the code
in `vim::surrounds::Vim.change_surrounds` so that it never adds or
removes whitespace when dealing with any type of quotes.
Closes#12247
Release Notes:
- Fixed whitespace handling when changing surrounding pairs to quotes in
vim mode
This PR fixes an issue where extension operations would never show in
the activity indicator despite this being implemented for ages. This
happened because we were always returning `None` whenever the app has a
global auto updater, which is always the case, so the code path for
showing extension updates in the indicator could never be hit despite
existing prior. Also slightly improves the messages shown for ongoing
extension operations, as these were previously context unaware.
While I was at this, I also quickly took a stab at cleaning up some
remotely related stuff, namely:
- The `AnimationExt` trait is now by default only implemented for
anything that also implements `IntoElement`. This prevents
`with_animation` from showing up for e.g. `u32` within the suggestions
(finally).
- Commonly used animations are now implemented in the
`CommonAnimationExt` trait within the `ui` crate so the needed code does
not always need to be copied and element IDs for the animations are
truly unique.
Relevant change here regarding the original issue is the change from the
`return match` to just a `match` within the activitiy indicator, which
solved the issue at hand.
If we find this to be too noisy at some point, we can easily revisit,
but I think this holds important enough information to be shown in the
activity indicator, especially whilst developing extensions.
Release Notes:
- Extension installation and updates will now be shown in the activity
indicator.
