# Objective
- Fixes#59654
When the same file was added as a mention more than once in the Agent
Panel,
the label would expand to the full absolute path instead of staying at
the
base filename.
Before
<img width="734" height="438" alt="Screenshot 2026-06-24 at 11 19 30 AM"
src="https://github.com/user-attachments/assets/94450eef-088a-49aa-97d3-80d6dd06925b"
/>
## Solution
The algorithm in `compute_disambiguation_details` treats two
entries with the same URI as a "collision" and tries to resolve it by
adding
more parent path components. Since the paths are identical it can never
succeed and keeps escalating to the full path.
Fix by deduplicating URIs via a `HashSet` before passing them to
`compute_disambiguation_details`, so identical mentions share a single
entry
and are never flagged as collisions.
## Testing
Reproduced by opening the Agent Panel, adding the same file as a mention
twice, and observing the label. Before: label showed the full absolute
path.
After: label stays at the base filename.
Tested on macOS.
After
<img width="638" height="215" alt="Screenshot 2026-06-24 at 11 20 09 AM"
src="https://github.com/user-attachments/assets/5acac2bd-d5a3-41c4-827c-84ccdc3a064d"
/>
## Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
---
Release Notes:
- Fixed duplicate file mentions in the Agent Panel showing the full
absolute
path instead of the base filename.
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
The width of the recent projects picker was set too small. When
selecting an item that is already opened an additional action (remove
current window) appears that overflows. This sets the width such that
that no longer occurs.
fixes: #59847
Release Notes:
- N/A (bug on preview only)
# Objective
Closes EP-78
- Describe the objective or issue this PR addresses.
- If you're fixing a specific issue, use "Fixes #X" for each issue as
[described in the GitHub
docs](https://docs.github.com/en/issues/tracking-your-work-with-issues/using-issues/linking-a-pull-request-to-an-issue#linking-a-pull-request-to-an-issue-using-a-keyword).
## Solution
- Describe the solution used to achieve the objective above.
## Testing
- Did you test these changes? If so, how?
- Are there any parts that need more testing?
- How can other people (reviewers) test your changes? Is there anything
specific they need to know?
- If relevant, what platforms did you test these changes on, and are
there any important ones you can't test?
## Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content adheres to Zed's UI standards
([UX/UI](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
and
[icon](https://github.com/zed-industries/zed/blob/main/crates/icons/README.md)
guidelines)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
## Showcase
> This section is optional. If this PR does not include a visual change
or does not add a new user-facing feature, you can delete this section.
- Help others understand the result of this PR by showcasing your
awesome work!
- If this PR includes a visual change, consider adding a screenshot,
GIF, or video
- A before/after comparison is very useful for changes to existing
features!
While a showcase should aim to be brief and digestible, you can use a
toggleable section to save space on longer showcases:
<details>
<summary>Click to view showcase</summary>
My super cool demos here
</details>
---
Release Notes:
- N/A or Added/Fixed/Improved ...
# Objective
Closes EP-9
- Describe the objective or issue this PR addresses.
- If you're fixing a specific issue, use "Fixes #X" for each issue as
[described in the GitHub
docs](https://docs.github.com/en/issues/tracking-your-work-with-issues/using-issues/linking-a-pull-request-to-an-issue#linking-a-pull-request-to-an-issue-using-a-keyword).
## Solution
- Describe the solution used to achieve the objective above.
## Testing
- Did you test these changes? If so, how?
- Are there any parts that need more testing?
- How can other people (reviewers) test your changes? Is there anything
specific they need to know?
- If relevant, what platforms did you test these changes on, and are
there any important ones you can't test?
## Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content adheres to Zed's UI standards
([UX/UI](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
and
[icon](https://github.com/zed-industries/zed/blob/main/crates/icons/README.md)
guidelines)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
## Showcase
> This section is optional. If this PR does not include a visual change
or does not add a new user-facing feature, you can delete this section.
- Help others understand the result of this PR by showcasing your
awesome work!
- If this PR includes a visual change, consider adding a screenshot,
GIF, or video
- A before/after comparison is very useful for changes to existing
features!
While a showcase should aim to be brief and digestible, you can use a
toggleable section to save space on longer showcases:
<details>
<summary>Click to view showcase</summary>
My super cool demos here
</details>
---
Release Notes:
- N/A or Added/Fixed/Improved ...
---------
Co-authored-by: Oleksiy Syvokon <oleksiy.syvokon@gmail.com>
Co-authored-by: Oleksiy Syvokon <oleksiy@zed.dev>
# Objective
- Polish the solo diff review UI by reducing redundant controls and
hiding inactive fold controls in split diff views.
## Solution
- Collapse the solo diff header's `Stage File` / `Unstage File` actions
into one button that shows the action relevant to the file's current
staging state.
- Stop reserving fold gutter space and rendering fold crease toggles for
editors with companion snapshots, matching the existing LHS split diff
behavior.
## Testing
- `cargo fmt --check`
- `cargo check -p git_ui`
## Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content adheres to Zed's UI standards
([UX/UI](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
and
[icon](https://github.com/zed-industries/zed/blob/main/crates/icons/README.md)
guidelines)
- [ ] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
## Showcase
Before, the solo diff header showed both file-level staging actions at
once, and the RHS split diff gutter could show fold arrows that did not
work in that view. After this change, the header shows the relevant
file-level staging action and split companion panes no longer show
inactive fold controls.
Release Notes:
- Improved solo diff controls by showing the relevant file staging
action and hiding inactive fold toggles in split diffs.
Resolves#56068
Remote extension sync previously only sent extensions that can run
remotely: language servers, debug adapters, and debug locators. That
excluded language-only extensions such as TOML, so a remote session
could have Tombi synced and loaded while the remote language registry
still lacked TOML's language definition.
This diff keeps the remote sync set narrow: it starts from extensions
that can run remotely, then adds the extension that defines each
language referenced by those extensions' language servers. This installs
the TOML language config alongside Tombi without syncing unrelated
language-only extensions, allowing project settings schemas to recognize
language names like TOML and allowing remote language servers to attach
to buffers whose language is provided by a separate extension.
Release Notes:
- Fixed remote sessions failing to recognize languages provided by
language-only extensions.
---------
Co-authored-by: Finn Evers <finn.evers@outlook.de>
# Objective
GPUI apps currently don't have a cross-platform way to react when the
system wakes from sleep.
## Solution
Add `Application::on_system_wake`, backed by platform hooks for macOS,
Windows, and Linux. The platform listeners are registered lazily once an
app installs a wake callback.
## Testing
- N/A
## Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content adheres to Zed's UI standards
([UX/UI](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
and
[icon](https://github.com/zed-industries/zed/blob/main/crates/icons/README.md)
guidelines)
- [ ] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Release Notes:
- N/A
## Summary
When you change a picker's preview position (toggle it, or set it
right/below/hidden), that choice should be remembered the next time the
picker opens. Previously it wasn't: the last-used layout was only
written to the database as a side effect of
`store_shape_for_this_layout`, which runs exclusively when a resize
completes (`render.rs`). So changing only the preview position without
also dragging to resize, was forgotten on the next open.
All layout changes funnel through a single chokepoint,
`set_preview_layout`, but it only updated the in-memory `preview.layout`
and reloaded the per-layout shape; it never persisted the layout key.
## Solution
`set_preview_layout` now persists the layout immediately via a new
`store_last_layout` helper, which writes just the
`LAST_PREVIEW_LAYOUT` key — deliberately without writing the current
shape, so a never-resized layout's default size isn't frozen into the
database as if the user had chosen it.
Both persistence paths are now correct:
- Resize completes — saves size and layout (unchanged).
- Position-only change (toggle / set right / set below / set hidden)
saves the layout right away.
This applies to every picker with a preview (file finder, text finder),
since the persistence is keyed generically by the picker delegate name.
## Testing
- Manual: open the file and text finder → switch the preview to Below
(or Right) without resizing → close and reopen → it returns in that
position; hide the preview → reopen → it stays hidden. Resizing still
persists size as before.
Release Notes:
- Fixed picker preview position not being remembered when changed
without resizing the picker
---------
Co-authored-by: ozacod <ozacod@users.noreply.github.com>
# Objective
Fixes https://github.com/zed-industries/zed/issues/59676
## Solution
Right now the queue / fetch logic is mixing async and sync flows. A task
might append something to the queue, but the queue fetching does not
respect/wait on that. Making this synchronous and in-memory fixes this.
`add_ssh_server` also appends it to the end of the queue, so we need to
get the last index, not index 0.
## Testing
I manually tested it before and after by reproducing the steps in the
original issue. Added a regression test afterwards.
## Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content adheres to Zed's UI standards
([UX/UI](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
and
[icon](https://github.com/zed-industries/zed/blob/main/crates/icons/README.md)
guidelines)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
---
Release Notes:
- N/A
---------
Co-authored-by: Yara <git@yara.blue>
We weren't handling these really before, just rendering the URI. This
should provide best-effort support for what we can render.
<img width="800" height="359" alt="image"
src="https://github.com/user-attachments/assets/51f101c7-c3be-4801-9000-62000a71faa6"
/>
(I will remove the extra labels that was just for testing)
Release Notes:
- acp: Support embedded resources in tool calls.
This removes sandbox prompt and documentation wording that claimed Git
metadata access exposes the inherited SSH agent socket. Git access
remains limited to protected metadata paths, and the Seatbelt wrapper no
longer carries SSH-agent-specific handling or examples.
Release Notes:
- Fixed AI sandboxing documentation to avoid implying Git metadata
access grants SSH agent access.
# Objective
Fixes#57850
NBSP, NNBSP, and NBH characters were causing breaks when text is wrapped
in editor when they should instead be non-breaking.
## Solution
I looked for an official spec and found [Unicode Line Breaking
Algorithm](https://www.unicode.org/reports/tr14/) which has these
characters listed under the non-breaking ("glue") section. I opted to
add these characters in manually within the logic that checks if a
character is a 'word' character or not, following the pattern for how
other groups of unicode characters are checked.
## Testing
- Tests were added under `test_is_word_char` and `test_wrap_line`
sections, and the expected behavior was verified by building zed and
using these characters in the editor.
- I don't think anything else needs testing now. In the future, it could
make sense to make sure the entire unicode spec is followed for line
breaking, as I imagine there are other obscure characters that could
have this same issue. (a `unicode-linebreak` crate is imported elsewhere
in the codebase but not used here)
- Changes can be tested by building this commit and following the
original instructions outlined in the related issue.
- Tested on linux
## Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments (**no unsafe
blocks**)
- [x] The content adheres to Zed's UI standards
([UX/UI](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
and
[icon](https://github.com/zed-industries/zed/blob/main/crates/icons/README.md)
guidelines)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
## Showcase
<details>
<summary>Before/After</summary>
### Before
<img width="449" height="443" alt="before"
src="https://github.com/user-attachments/assets/4d438966-1bab-46e4-a018-cd648395a1ca"
/>
### After
<img width="448" height="445" alt="after"
src="https://github.com/user-attachments/assets/c7ce7166-9c22-4083-850f-8d8acfc42b56"
/>
</details>
---
Release Notes:
- N/A
## Bug
Before switching macOS appearance:
<img width="154" height="39" alt="before switching OS theme"
src="https://github.com/user-attachments/assets/2444fafe-21a7-47f6-84ac-ec33a61a6523"
/>
After switching macOS appearance:
<img width="167" height="36" alt="after switching OS theme"
src="https://github.com/user-attachments/assets/d4a34555-2964-4e92-8d6a-24da7620fc49"
/>
After focusing the window following the macOS appearance switch:
<img width="151" height="37" alt="after focusing the window after
switching the OS theme"
src="https://github.com/user-attachments/assets/6bc2e996-256d-448d-a11b-fdc2f0e01f58"
/>
## Fix
Reapply GPUI's custom macOS traffic-light button position after AppKit
reports an effective appearance change and after GPUI's appearance
observers have been notified.
## Cause
GPUI stores a custom `traffic_light_position` for windows that use
transparent/custom titlebars. When macOS switches Light/Dark appearance,
AppKit can relayout the standard window buttons, which resets the
traffic lights to AppKit's default vertical position. The next normal
window focus/layout pass moves them back, causing a visible jump.
This change keeps the existing appearance observer callback behavior,
then re-runs `move_traffic_light()` after the callback has completed.
## Validation
- `cargo check -p gpui_macos --release --features runtime_shaders`
- Manually verified with a GPUI window using a transparent titlebar,
custom `traffic_light_position`, and an appearance observer that
refreshes the window.
- Toggled macOS Light/Dark appearance while the window was visible;
traffic lights no longer shifted upward.
Release Notes:
- N/A
## Summary
When opening the text finder (`text_finder::Toggle`), pre-populate its
query the same way project search already seeds itself. Previously
`seed_query` only asked the active editor for the word under the cursor,
so opening the text finder from a project search tab or a focused buffer
search bar started empty.
`seed_query` now tries these sources in priority order:
1. **Active project search** — its current query
(`ProjectSearchView::search_query_text`).
2. **Focused buffer search bar** — its query, reusing project search's
existing `buffer_search_query` helper (promoted from private to
`pub(crate)` so there's one source of truth rather than a copy).
3. **Word under the cursor** — the existing `query_suggestion` fallback,
still honoring the `seed_search_query_from_cursor` setting.
This mirrors how project search is fed from the buffer search bar
(`existing_or_new_search`), keeping the seeding behavior consistent
across the three search surfaces.
## Testing
- Manual: open project search with a query → toggle text finder → query
carries over; cmd+f with a query focused → toggle text finder → query
carries over; otherwise the word under the cursor seeds it as before.
Release Notes:
- Improved text finder to pre-fill its query from the active project
search or buffer search
---------
Co-authored-by: ozacod <ozacod@users.noreply.github.com>
Co-authored-by: Yara 🏳️⚧️ <git@yara.blue>
## Data table: Add left inset for scrollbar alignment with pinned
columns
When pinned columns are enabled in the data table, the horizontal
scrollbar previously extended across the full table width, including
behind the pinned columns. This change adds a left inset to the
scrollbar so it only spans the scrollable area, properly aligning with
the content layout.
## Demo (look at bottom scrollbar)
| Before | After|
| --- | --- |
| <img width="1614" height="482" alt="image_2026-05-20_12-29-22"
src="https://github.com/user-attachments/assets/87050173-0f3a-46dd-a7e1-10f769c7b5b8"
/> | <img width="1526" height="420" alt="image_2026-05-20_12-32-48"
src="https://github.com/user-attachments/assets/5041069f-2289-449b-8d30-8bcb3d0890eb"
/> |
### Changes
- Add `pinned_width()` and `scrollable_width()` helper methods to
`ResizableColumnsState` to compute column widths
- Refactor scrollbar rendering to use absolute positioning with inset
when pinned columns are present
- Scrollbar now correctly positions itself to span only the scrollable
(non-pinned) column area
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Release Notes:
- Improved horizontal scrollbar alignment when pinned columns are active
- N/A or Added/Fixed/Improved ...
---------
Co-authored-by: Smit Barmase <heysmitbarmase@gmail.com>
## Summary
Inline `<br>` tags in markdown (e.g. `<br>`, `<br/>`, `<br />`) were
rendered as literal text instead of line breaks when images are around.
VS Code and GitHub both treat them as hard line breaks.
The fix has two layers:
**Parser (`parser.rs`):** The text-merge loop was consuming `InlineHtml`
events as empty text. `<br>` variants are now excluded from that loop
and converted to `HardBreak` events instead, matching the semantics of a
trailing two-space or backslash line break in CommonMark.
**Renderer (`markdown.rs`):** Image paragraphs use a flex row layout, so
text `\n` has no effect. `HardBreak` inside an image paragraph now
inserts a `div().w_full().h_0()` child, which forces a flex-wrap row
break. Images also get `mr_2()`/`mb_2()` margins for spacing, replacing
the previous `SoftBreak`-as-space hack. `SoftBreak` is suppressed inside
image paragraphs to avoid phantom whitespace elements disrupting the
flex layout.
## Screenshots
### Before
How Zed WAS rendering this:
<img width="629" height="601" alt="Screenshot 2026-04-17 at 07 53 26"
src="https://github.com/user-attachments/assets/cbfa6c81-83ea-467a-b73f-a2efef953bb4"
/>
### Comparison
This is how Github renders this:
<img width="1006" height="911" alt="Screenshot 2026-04-17 at 07 06 49"
src="https://github.com/user-attachments/assets/6ee286c6-026b-4a85-a931-c61cf24370ba"
/>
This is how VS Code renders this:
<img width="1511" height="947" alt="Screenshot 2026-04-17 at 07 03 36"
src="https://github.com/user-attachments/assets/1bc41c85-bb3d-48c6-9e21-cd96327c43c6"
/>
### After
This is the end result after the changes:
<img width="821" height="807" alt="Screenshot 2026-04-17 at 07 45 32"
src="https://github.com/user-attachments/assets/8882abaa-dc07-45fb-9aaa-a908712dce65"
/>
[Created this Github repo for testing
](https://github.com/davidalecrim1/zed-md-test)
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Release Notes:
- Fixed inline `<br>` tags in markdown previews rendering as literal
text instead of line breaks
---------
Co-authored-by: Smit Barmase <heysmitbarmase@gmail.com>
The issue here was that `end_slot_on_hover` would currently only work if
`end_slot` was also set. With this, it will now work even if there is no
`end_slot`.
Release Notes:
- N/A
# Objective
Consider a simple Python file with the following content, where both a
method and a function are defined:
```python
class TestCase:
def a_test_method(self): # a method
print("This is a method test")
def a_test_function(): # a function
print("This is a function test")
```
Below this, if we type `a_test_func` and trigger language server
completion, we would expect to see `a_test_function` in the completion
list. Similarly, if we have:
```python
test = TestCase()
test.a_testˇ
```
and trigger language server completion, we should see `a_test_method` in
the completion list.
In both cases, we expect the items in the completion list to have the
same color highlighting as they do in the editor buffer, but currently,
they do not. This is because Zed uses the wrong queries to retrieve the
highlight. Taking the `ty` LSP adapter as an example:
https://github.com/zed-industries/zed/blob/ef7fc3c28828ff9edbe35849567634c13ce2ee3b/crates/languages/src/python.rs#L323-L332
the key `"function.method"` and `"function"` are used to look up the
method and function highlight IDs, but these keys do not exist in
`crates/grammars/src/python/highlights.scm`, which defines the syntax
highlighting. The correct keys here should be `"function.method.call"`
and `"function.call"`.
## Solution
- As stated above, change the keys from `"function.method"` and
`"function"` to `"function.method.call"` and `"function.call"`,
respectively.
- Additionally, since we have several built-in language servers for
Python, a minor refactoring has been done to reduce duplicate code and
simplify future changes.
## Testing
- I have tested this change locally; the before and after comparison is
provided below.
## Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content adheres to Zed's UI standards
([UX/UI](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
and
[icon](https://github.com/zed-industries/zed/blob/main/crates/icons/README.md)
guidelines)
- [ ] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
## Showcase
<details>
<summary>Click to view showcase</summary>
| Before | After |
| :---: | :---: |
| <img width="767" height="308" alt="function_before"
src="https://github.com/user-attachments/assets/943852d6-f54e-411d-916f-add7aac57956"
/> | <img width="767" height="308" alt="function_after"
src="https://github.com/user-attachments/assets/3a817229-ce32-4d13-88b8-9e9151869338"
/> |
| <img width="767" height="308" alt="method_before"
src="https://github.com/user-attachments/assets/cbc0f4f4-be6e-44d3-9798-df29df5731a3"
/> | <img width="767" height="308" alt="method_after"
src="https://github.com/user-attachments/assets/6e8f7e84-c5c6-4b37-a2c5-10820f12a09d"
/> |
</details>
---
Release Notes:
- Improved Python completion item highlighting for methods and
functions.
Agent terminal commands that request Git metadata access now verify
linked-worktree and submodule metadata before moving Git paths from
protected to writable sandbox paths. This prevents a mutated `.git`
gitfile from redirecting a grant to unrelated repository metadata while
preserving standard submodule layouts.
Release Notes:
- Improved agent terminal sandboxing for Git metadata access.
---------
Co-authored-by: Martin Ye <martin@zed.dev>
Closes https://github.com/zed-industries/zed/issues/52576
## Overview
Adds prompt caching support for Anthropic Claude models when accessed
via OpenRouter.
This mirrors Zed's native Anthropic provider by using explicit per-block
`cache_control` breakpoints:
- A long-lived 1-hour breakpoint on the system message's last text
block, covering the tools and system prefix.
- A default 5-minute breakpoint on the last `cache: true` conversation
message.
The implementation intentionally avoids OpenRouter's top-level automatic
`cache_control` field so routing remains available across
Anthropic-compatible providers including Anthropic, Bedrock, and Vertex
AI. It also adds `session_id` sticky routing from the request thread ID
and maps OpenRouter cache usage fields into Zed's token usage
accounting.
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Release Notes:
- Fixed Anthropic prompt caching when using OpenRouter.
---------
Co-authored-by: Anant Goel <anant@zed.dev>
# Objective
Fix OpenAI-compatible reasoning/thinking support. Previously,
`OpenAiCompatibleLanguageModel` never reported thinking support, so
configured `reasoning_effort` values did not reliably reach the wire and
Responses API reasoning state could miss `include:
["reasoning.encrypted_content"]`.
Fixes#58289
Addresses #59207 for OpenAI-compatible Responses models configured with
`reasoning_effort`.
## Solution
- Treat a non-`none` configured `reasoning_effort` as OpenAI-compatible
thinking support.
- Expose common OpenAI-style effort levels in the agent UI, using the
configured effort as the default.
- Honor selected reasoning effort for OpenAI-compatible chat-completions
requests.
- Send `reasoning_effort: "none"` when thinking is disabled for a
configured reasoning model.
- Preserve native OpenAI behavior by continuing to use native
model-specific supported efforts and `max_completion_tokens`.
- Add an OpenAI-compatible `max_tokens_parameter` capability for
endpoints that expect output limits as `max_tokens`.
- Parse common streamed thinking fields (`reasoning` and
`reasoning_content`).
- Add provider setup UI and docs for configuring OpenAI-compatible
reasoning models.
## Testing
- `cargo fmt -p language_model_core -p agent_ui -p language_models`
- `cargo test -p language_models provider::open_ai_compatible::tests
--lib`
- `cargo test -p language_models provider::open_ai::tests --lib`
- `cargo test -p open_ai completion::tests --lib`
- `cargo test -p agent_ui
agent_configuration::add_llm_provider_modal::tests --lib`
- `git --no-pager diff --check`
## Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content adheres to Zed's UI standards
([UX/UI](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
and
[icon](https://github.com/zed-industries/zed/blob/main/crates/icons/README.md)
guidelines)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
---
Release Notes:
- Improved OpenAI-compatible provider setup for reasoning models.
---------
Co-authored-by: Anant Goel <anant@zed.dev>
Summary:
- Add a shell-level startup handshake owned by the terminal.
- Send a split startup marker through the shell and run the configured
init command after the marker is observed, preventing local echo from
satisfying the handshake.
- Keep display-only/restored terminal behavior unchanged while
preserving startup output when shells fail or users type first.
Tests:
- `cargo test -p terminal test_init_command_startup_marker --
--nocapture`
- `cargo test -p terminal write_init_command_after_startup --
--nocapture`
- `cargo test -p agent_ui init_command -- --nocapture`
- `./script/clippy -p terminal -p agent_ui`
Closes AI-426
Release Notes:
- Fixed terminal init commands sometimes appearing or running
incorrectly when using custom shells
---------
Co-authored-by: Anant Goel <anant@zed.dev>
## Objective
Make text finder match project search and buffer search by
pre-populating its query from the active editor according to
`seed_search_query_from_cursor`.
## Solution
- Seed text finder from the active editor's query suggestion when
opening it.
- Pass the seeded query into the picker and select the query text so
typing replaces it immediately.
## Testing
- Manually tested visually by opening text finder from an editor with a
selection and with the cursor on text, confirming the query is
pre-populated and selected.
## Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Release Notes:
- Improved text finder to seed its query from the active editor.
Co-authored-by: ozacod <ozacod@users.noreply.github.com>
Co-authored-by: Yara 🏳️⚧️ <git@yara.blue>
Agent terminal sandboxing now protects `.git` metadata by default and
exposes an explicit `allow_git_access` approval. Without it, file
contents of the `.git` directories for opened worktrees and discovered
repositories (including a linked worktree's common `.git`) cannot be
read or written, though their metadata stays visible; when approved,
those Git directories become writable so commands like fetch/commit
work. SSH commit signing keeps working because the inherited
`SSH_AUTH_SOCK` is allowed as local Unix-socket IPC, which does not let
sandboxed commands send network packets to other machines. The Seatbelt
profile also allows PTY terminal-control ioctls so signing/passphrase
prompts can manage terminal state.
This intentionally restricts only `.git` itself (whose location we know
exactly, including the worktree case) rather than
`.gitignore`/`.gitattributes`/`.gitmodules`, since those can be nested
arbitrarily and the goal is to keep the policy expressible as a plain
deny-by-default allowlist that will port to Linux and Windows sandboxes
later.
Closes AI-334
Release Notes:
- Improved agent terminal sandboxing for Git metadata, Git worktrees,
and SSH commit signing.
Release Notes:
- Agent: Fixed a bug where trying to open the profile configuration
modal through the picker in the agent panel's message editor wouldn't
work through the keybinding.
Share marked selections across rendered project panel entries and defer
drag-related path cloning until drag and drop rendering is enabled. Also
defer symlink path string conversion until tooltip creation.
Release Notes:
- N/A or Added/Fixed/Improved ...
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [ ] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Release Notes:
- Added initial coverage set of Helix debugger keybindings.
# Objective
- Keep the Git panel view options menu open when changing view, sort, or
group options.
- Make the open menu reflect the selected option immediately, including
showing the Sort By section as soon as List view is selected.
## Solution
- Use a persistent context menu for the Git panel view options menu so
selecting an item rebuilds the menu instead of dismissing it.
- Track the open menu's view options state locally with
`GitPanelViewOptionsMenuState` so checkmarks and conditional sections
update before the settings file change is observed.
- Continue dispatching the existing Git panel actions so persisted
settings and panel updates keep using the current code path.
## Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content adheres to Zed's UI standards
([UX/UI](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
and
[icon](https://github.com/zed-industries/zed/blob/main/crates/icons/README.md)
guidelines)
- [ ] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
## Showcase
- N/A
---
Release Notes:
- Fixed Git Panel view options menu selections updating only after
reopening the menu.
While exploring the `vim` crate, I found that `ShellCommand` is
registered twice with identical closure bodies in
`crates/vim/src/command.rs`
[L331-L357](https://github.com/zed-industries/zed/blob/main/crates/vim/src/command.rs#L331-L357)
. This PR removes the redundant code block to clean up.
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Release Notes:
- N/A
Co-authored-by: dino <dinojoaocosta@gmail.com>
The `go-subtest` task template wrapped the `-run` arg in single quotes
for shell safety. This works for Run Test (terminal strips quotes), but
Debug Test sends the arg through Delve’s DAP protocol with no shell
involved, so the literal quote characters ended up in the regex and
prevented any match.
All other Go task templates (`go-test`, `go-testify-suite`,
`go-table-test-case`) use the backslash-escaped format which `GoLocator`
already knows how to handle. Align the `go-subtest` template to the same
format.
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [ ] Unsafe blocks (if any) have justifying comments
- [ ] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable
Closes#53230.
Release Notes:
- Fixed Debug Test for Go subtests
---------
Co-authored-by: Smit Barmase <heysmitbarmase@gmail.com>
Pulls in https://github.com/zed-industries/livekit-rust-sdks/pull/13 to
fix a cause of panics with webrtc
Fixes ZED-5J7
Release Notes:
- N/A or Added/Fixed/Improved ...
The recent "pickers with previews" overhaul left the picker
sizing/presentation API spread across several overlapping knobs that
every call site had to set correctly — and many didn't, causing pickers
across the app to render at the wrong width, lose their container, or
stop dismissing. This PR consolidates that surface into a small,
hard-to-misuse API and makes correct sizing the default.
Net effect: a plain `Picker::uniform_list(delegate, …)` now renders
correctly out of the box (standard width, standard max-height, shrinks
to fit, dismisses properly), and the ~35 call sites only specify what
genuinely differs.
## API changes
**Presentation** — three overlapping booleans (`is_modal`, `is_popover`,
`is_resizable`) collapsed into one enum, with resizability living inside
the only variant where it's meaningful:
```rust
enum Presentation {
Modal { resizable: bool }, // own chrome, dismisses on blur, optionally resizable
Popover, // own chrome, dismisses on blur, never resizable
Embedded, // host container owns chrome + dismissal
}
```
- `modal(bool)` is **removed** in favor of explicit, self-documenting
builders:
- *(default)* → `Modal` (resizable iff it has a preview)
- `.popover()` → `Popover` (menu-attached surfaces)
- `.embedded()` → `Embedded` (pickers nested in a larger modal/view)
- Dynamic callers use `.when(cond, Picker::embedded)` (added `impl
FluentBuilder for Picker`).
**Sizing** — preview-vs-not now drives everything; the manual padding
knob is gone:
| Before | After |
|---|---|
| `vertical_padding` field + `no_vertical_padding()` | removed — derived
from whether a preview is visible |
| `height(...)` (ambiguous: fixed vs max) | `max_height(...)` (plain
pickers shrink-to-fit, capped here) |
| `minimum_results_width(...)` | removed — a plain picker's min width
tracks its opening width; preview pickers use standard internal pane
mins |
| default size = 60% viewport | default = `DEFAULT_MODAL_WIDTH` (34rem)
× `DEFAULT_MODAL_MAX_HEIGHT` (24rem, max) |
| resize handles gated on `is_modal` | gated on `is_resizable` (new
`resizable(bool)` builder; auto-`true` for preview pickers) |
Call sites now only override the exceptions: narrow popover selectors
(`initial_width`), the taller outline view (`max_height`), and preview
pickers (constructed via `*_with_preview`).
## Behavior fixes
- **Wrong widths everywhere**: pickers were falling back to
60%-of-viewport because the original migration set
`minimum_results_width` but never `initial_width`. Fixed at the source
via the new defaults.
- **Popovers had no container and wouldn't dismiss**: `is_modal=false`
was suppressing both the elevated background *and* blur-dismiss. Split
out so popovers keep their chrome and dismiss on click-away/escape. This
fixed the agent-panel model/profile selectors, sidebar recent projects,
and the settings theme/font/icon/ollama pickers (which were incorrectly
using `modal(false)`).
- **Sidebar recent projects stretched to full height**: was missing the
shrink-to-fit behavior; now capped and content-sized like other
popovers.
- **Preview crash**: removed an over-strict `debug_assert!` that
panicked when previewing an empty file (`message == None && editor
empty` is valid).
- **Preview-aware default size**: pickers open at standard width with
the preview hidden, and expand to the larger "telescope" size when a
preview is shown. Fixes the text finder rendering super-wide by default,
and makes the file finder expand (rather than cram its results) when you
toggle the preview.
---
Release Notes:
- N/A
## Summary
Fixes#37815.
`State::fetch_models()` calls `/api/tags` to list models, then calls
`/api/show` for **every** model in that list to get its capabilities,
collecting the results with `collect::<Result<Vec<_>>>()?`. If
`/api/show` errors for even one model, the whole batch fails and
`fetched_models` is never populated. Since `is_authenticated()` is
defined as `!self.fetched_models.is_empty()`, this means a single bad
model permanently breaks both authentication state and the model picker
for the entire Ollama provider - with no error surfaced anywhere (not
the UI, not the logs), which matches the reports in #37815 of "Connect
does nothing" / "no logs, no nothing".
I hit this myself: I had a stale local reference to an Ollama Cloud
model that had been retired server-side. `/api/show` for that one model
returned `410 Gone`, which silently broke Connect and the model picker
for every other model too. Removing the retired model with `ollama rm`
fixed it immediately, which confirmed the root cause.
## Fix
Instead of aborting the whole fetch on the first error, skip individual
models that fail `/api/show` and log a warning, keeping the rest.
Extracted this into a small `skip_failed_models` helper so it's
unit-testable without mocking HTTP.
## Disclosure
I used Claude (Anthropic's Claude Code) to help track down this root
cause (tracing through `fetch_models`/`is_authenticated` in this file)
and draft the fix + tests below. I reviewed and understand the change -
it's a small, targeted fix to a single function plus two unit tests for
the new helper.
## Test plan
- [x] `cargo check -p language_models` passes
- [x] `cargo test -p language_models --lib ollama::` - all 3 tests pass
(the 2 new ones plus the existing
`test_merge_settings_preserves_display_names_for_similar_models`,
unaffected by this change)
- [x] `cargo fmt -p language_models -- --check` - no diff
## Release Notes
Release Notes:
- Fixed Ollama models silently failing to show up in the model picker
(and "Connect" appearing to do nothing) when a single model's details
couldn't be fetched, e.g. a retired Ollama Cloud model
---------
Co-authored-by: MrSubidubi <finn@zed.dev>
Co-authored-by: Ben Brandt <benjamin.j.brandt@gmail.com>
- **Revert "livekit: Preserve tokens on channel rejoin (#59388)"**
- **Revert "call: Log LiveKit connection info refresh outcomes in retry
loop (#59205)"**
- **Revert "audio: Fix phantom presence in channels (#59195)"**
- **proto: Reserve RejoinRoomResponse field 4 after revert**
We've observed a spike of phantom collaborator issues after this fixes.
This sucks and I'm quite unhappy about it.
# Objective
- Describe the objective or issue this PR addresses.
- If you're fixing a specific issue, use "Fixes #X" for each issue as
[described in the GitHub
docs](https://docs.github.com/en/issues/tracking-your-work-with-issues/using-issues/linking-a-pull-request-to-an-issue#linking-a-pull-request-to-an-issue-using-a-keyword).
## Solution
- Describe the solution used to achieve the objective above.
## Testing
- Did you test these changes? If so, how?
- Are there any parts that need more testing?
- How can other people (reviewers) test your changes? Is there anything
specific they need to know?
- If relevant, what platforms did you test these changes on, and are
there any important ones you can't test?
## Self-Review Checklist:
- [ ] I've reviewed my own diff for quality, security, and reliability
- [ ] Unsafe blocks (if any) have justifying comments
- [ ] The content adheres to Zed's UI standards
([UX/UI](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
and
[icon](https://github.com/zed-industries/zed/blob/main/crates/icons/README.md)
guidelines)
- [ ] Tests cover the new/changed behavior
- [ ] Performance impact has been considered and is acceptable
## Showcase
> This section is optional. If this PR does not include a visual change
or does not add a new user-facing feature, you can delete this section.
- Help others understand the result of this PR by showcasing your
awesome work!
- If this PR includes a visual change, consider adding a screenshot,
GIF, or video
- A before/after comparison is very useful for changes to existing
features!
While a showcase should aim to be brief and digestible, you can use a
toggleable section to save space on longer showcases:
<details>
<summary>Click to view showcase</summary>
My super cool demos here
</details>
---
Release Notes:
- N/A
---------
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
# Objective
The reasoning for this is that it is currently rather hard to find the
option to add a remote MCP in the app - there is no hint that this is
supported and (without using the command palette) requires up to 5
clicks before the proper modal is shown.
The idea for this here is to make adding a remote server the default and
also make it more visibile throughout the app - the `AddContextServer`
action will now default to remote and this also adds two context menu
entries to directly open the modal with the remote option preselected.
Open for feedback - I admit this might not be the most optiomal
solution, but at least it surfaces the availablilty better and
(hopefully) makes adding these easier.
Release Notes:
- Added an entry to the agent panel options menu to quickly add a remote
MCP server
# Objective
Telemetry events generated **on a remote server** were silently dropped,
and all remote telemetry that *was* reported got attributed to the local
client's OS.
## Solution
This PR makes server-originated events flow to the telemetry pipeline
and attributes them to the actual remote host (connection type, OS,
version, architecture), plus adds a transport-agnostic connection event.
## Additional Notes
This removes the "SSH Project Opened" event and replaces it with a
"Remote Connection Established" one that has more structured info and is
also more useful in where it gets invoked.
The server unconditionally sends back telemetry to the client, the
client is then responsible for filtering on whether telemetry is enabled
by the user or not as the server does not have knowledge of those
settings itself.
Release Notes:
- N/A
