I noticed that the minimap seems to render underlines with the same
thickness as the main editor, which looks a bit off. This becomes much
more noticeable when enabling `semantic_token_rules` (due to the
increased number of underlines):
```json
"global_lsp_settings": {
"semantic_token_rules": [
{
"token_modifiers": ["mutable"],
"underline": true,
},
],
}
```
Looking at the existing code, I found that diagnostic underlines already
check `editor_style.show_underlines` to ensure they are only displayed
in the main editor. To maintain consistency, I applied the same
filtering logic to `chunk_highlight` so that these underlines are no
longer rendered in the minimap.
Before：
<img alt="CleanShot 2026-02-06 at 02 28 31@2x"
src="https://github.com/user-attachments/assets/16401154-23f5-43ef-a7a8-b1035c19e076"
/>
After：
<img alt="CleanShot 2026-02-06 at 02 31 36@2x"
src="https://github.com/user-attachments/assets/979a04be-e585-44be-9c42-4dfab7b89186"
/>
Release Notes:
- N/A *or* Added/Fixed/Improved ...
Since the read times always correspond to an action log call anyway, we
can let the action log track this internally, and we don't have to
provide a reference to the Thread in as many tools.
Release Notes:
- N/A
---------
Co-authored-by: Bennet Bo Fenner <bennetbo@gmx.de>
Co-authored-by: MrSubidubi <dev@bahn.sh>
`.cppm` is the widely used extension for C++20 module interface units,
supported by MSVC, Clang, and GCC. Currently Zed doesn't recognize it as
C++, so users get no syntax highlighting or LSP support.
Changes:
`crates/languages/src/cpp/config.toml`: add cppm to path_suffixes
`crates/theme/src/icon_theme.rs`: add cppm to the C++ icon matcher
https://github.com/search?q=path%3A*.cppm&type=code
Release Notes:
- N/A
Adds the actual tool name so people can turn it off if they want.
Release Notes:
- N/A
Co-authored-by: Bennet Bo Fenner <bennetbo@gmx.de>
Co-authored-by: MrSubidubi <dev@bahn.sh>
Previously, there was no way to view the full thread context
Release Notes:
- N/A
Co-authored-by: Bennet Bo Fenner <bennetbo@gmx.de>
Co-authored-by: MrSubidubi <dev@bahn.sh>
IAM users cannot be assumed; only IAM roles can be.
Before you mark this PR as ready for review, make sure that you have:
- [ ] Added a solid test coverage and/or screenshots from doing manual
testing
- [ ] Done a self-review taking into account security and performance
aspects
- [ ] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
Release Notes:
- N/A
Closes#46767
## Summary
The "Reveal in File Manager" action was shown in the context menu for
WSL projects (guarded by `is_via_wsl_with_host_interop`), but the action
handler in `Render` was only registered when `project.is_local()` —
which returns `false` for WSL. Dispatching the action without a handler
caused a crash.
Adds the same `is_via_wsl_with_host_interop(cx)` check to the handler
registration.
## Testing
- Ran `cargo test -p project_panel` — 78 passed, 0 failed
- Manual testing: connected to WSL Ubuntu, right-clicked a file in the
project panel, used "Reveal in File Manager" — Windows Explorer opened
correctly without crashing
Before you mark this PR as ready for review, make sure that you have:
- [x] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
- [x] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
Release Notes:
- Fixed a crash when using "Reveal in File Manager" on files in WSL
projects (#46767).
Continues #47114
Release Notes:
- Fixed REPL output width clamping to apply to the content area so
images don’t get clipped by controls
---------
Co-authored-by: MrSubidubi <finn@zed.dev>
Hi! The `...` entry in the `language_servers` setting was only explained
in a single bullet point, which led users to misconfigure their setup,
particularly when overriding defaults that disable certain servers with
`!`.
Add a detailed explanation of how `...` works and a table of examples
using Ruby's real server configuration to illustrate the override
behavior.
Before you mark this PR as ready for review, make sure that you have:
- [ ] Added a solid test coverage and/or screenshots from doing manual
testing
- [ ] Done a self-review taking into account security and performance
aspects
- [ ] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
Release Notes:
- N/A
The purpose of `register_available_lsp_adapter()` is to allow language
servers to be reused across multiple languages. Since adapters like
`ty`, `pylsp`, and `pyright` are specific to Python, there is no need to
register them for other languages. Additionally, registering them
directly to the global `LanguageRegistry` results in negligible resource
consumption.
We can then use the default settings to control the default language
server for Python, as referenced here:
https://github.com/zed-industries/zed/blob/2cc3d382bb669629471886c2ea9b43311c525099/assets/settings/default.json#L2119-L2130
Additionally, the documentation for Python has been updated to clarify
that the `"..."` syntax does not mean "keep the rest at default," but
rather "include all other available servers."
Before you mark this PR as ready for review, make sure that you have:
- [ ] Added a solid test coverage and/or screenshots from doing manual
testing (no sure how to add test for this)
- [x] Done a self-review taking into account security and performance
aspects
- [x] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
Release Notes:
- N/A *or* Added/Fixed/Improved ...
The main intention behind this change is to support uploading these
files
to GitHub. `.miniprof` is not a supported extension by GitHub, but
`.json` is.
The only “downside” to this change is that the cleanup process will have
to look
for `.miniprof` files AND `.miniprof.json` files. Maybe we can remove
that change
at a later date?
Ref:
https://docs.github.com/en/get-started/writing-on-github/working-with-advanced-formatting/attaching-files
Before you mark this PR as ready for review, make sure that you have:
- [ ] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
- [x] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
Release Notes:
- Changed miniprofiler file extension to `.miniprof.json`
Since we were no longer just returning a string, we need to update the
content in both success and error modes to get a nice rendering
experience.
Release Notes:
- N/A
Closes #ISSUE
Before you mark this PR as ready for review, make sure that you have:
- [ ] Added a solid test coverage and/or screenshots from doing manual
testing
- [ ] Done a self-review taking into account security and performance
aspects
- [ ] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
Release Notes:
- N/A *or* Added/Fixed/Improved ...
Closes #ISSUE
Before you mark this PR as ready for review, make sure that you have:
- [ ] Added a solid test coverage and/or screenshots from doing manual
testing
- [ ] Done a self-review taking into account security and performance
aspects
- [ ] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
Release Notes:
- Added support for providing an API key to OpenAI-compatible edit
prediction providers
When a downstream project was disconnected from the host (e.g. the guest
left the call), `disconnected_from_host_internal` did not clear
`client_subscriptions`. These subscriptions hold entries in the
`Client`'s entity subscription map, so a subsequent
`join_remote_project` with the same project ID would fail with "already
subscribed to entity".
The fix adds `self.client_subscriptions.clear()` to
`disconnected_from_host_internal`, matching what `unshare_internal`
already does for the host side.
Before you mark this PR as ready for review, make sure that you have:
- [x] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
Release Notes:
- collab: Fix unable to rejoin project bug ("already subscribed to
entity")
This can be used like `ep predict --provider
baseten:V0131GitMergeMarkersPrefix`. Since it doesn't require
load_project, it can be used with captured requests.
Release Notes:
- N/A
The main issue is that we weren't forwarding the proto messages through
the collab server to the host. After fixing that I added integration
tests to cover local worktrees, remote worktrees, and ssh worktrees.
I also fixed a bug with FakeRepository where it wouldn't name its
current branch as a worktree when calling git worktree, which doesn't
match the behavior of the git binary.
Before you mark this PR as ready for review, make sure that you have:
- [x] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
Release Notes:
- git: Fix bug that caused the git worktree picker from displaying and
creating worktrees over collab
I've been using
https://github.com/zed-industries/zed/issues/15046#issuecomment-3259286451
for half a year now, and it seems worthy of inclusion in the default
keymap.
Before you mark this PR as ready for review, make sure that you have:
- [ ] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
- [x] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
Release Notes:
- Added `ctrl-enter` keybind on macOS to type a newline in search bars
Release Notes:
- Windows: make auto-update more robust in the face of apps holding the
Zed.exe handle
---------
Co-authored-by: Jakub Konka <kubkon@jakubkonka.com>
We would mark `FileNotFound` as success and progress the update loop
which does not make much sense.
Release Notes:
- Do not skip update roll back in presence of FileNotFound errors on Windows.
Co-authored-by: Miguel Raz Guzman Macedo <raz@zed.dev>
## Problem
- On Linux, non-recursive watcher registrations remained path-cached
after deleting and recreating a directory in the same session.
- The recreated directory was not re-watched, so newly created child
entries under that path could be missing.
## Summary
- Remove directory watcher registrations when worktree paths are removed
from snapshot state.
- Ensure recreated directories can be watched again on Linux by allowing
`scan_dir` to re-add fresh watches.
- Add a Linux integration regression test for directory delete/recreate
path reuse and child file creation.
## Testing
- `cargo test -p project --features test-support --test integration
test_recreated_directory_receives_child_events -- --exact`
- `cargo test -p project --features test-support --test integration
test_rescan_and_remote_updates -- --exact`
## Related
- #46709
Release Notes:
- Fixed Linux worktree file watching so child entries appear after
deleting and recreating a directory at the same path.
### Summary
Makes file mention chips in the AI chat input clickable to open the
referenced files. Previously, chips like `@README.md` were purely visual
indicators with no interaction.
### Changes
- **Clickable mention chips**: Users can now click on file mentions in
the chat input to open those files in the editor
- **Support for all mention types**:
- Files → Opens in editor
- Files with line numbers → Opens and scrolls to line
- Directories → Reveals in project panel
- Threads → Navigates to thread
- Rules → Opens rules library
- URLs → Opens in browser
- **Handles files outside workspace**: Falls back to `open_abs_path()`
for files not in the current workspace
### Implementation
Threads `MentionUri` and `WeakEntity<Workspace>` through the crease
rendering pipeline:
1. Updated `insert_crease_for_mention()` to accept mention URI and
workspace references
2. Added click handler to `MentionCrease` component using `.when()` for
conditional attachment
3. Implemented file opening helpers that mirror the existing
`thread_view.rs::open_link()` logic
### Demo
https://github.com/user-attachments/assets/21b2afb7-7a86-4a0a-aba1-e24bb1b650c2
### Testing
Manually tested:
- [x] Clicking `@README.md` opens file
- [x] Clicking file with line numbers navigates correctly
- [x] Clicking directory reveals in project panel
- [x] Files outside workspace open via absolute path
### Files Changed
- `crates/agent_ui/src/mention_set.rs` - Thread URI/workspace through
pipeline
- `crates/agent_ui/src/ui/mention_crease.rs` - Add click handler and
file opening logic
- `crates/agent_ui/src/acp/message_editor.rs` - Update call sites
### Review feedback addressed
- Replaced `.when()` + `unwrap()` with `.when_some()` + `Option::zip()`
(`0e36efb4eb`)
- De-duplicated `open_file` and `open_file_at_line` into a single
function with `Option<RangeInclusive<u32>>` (`dbcbb69a4b`)
- Rebased onto latest `main` and resolved conflicts
Also update item 2 under Implementation from:
_Added click handler to MentionCrease component using `.when()` for
conditional attachment_
to:
_Added click handler to MentionCrease component using `.when_some()`
with `Option::zip()` for conditional attachment_
### Release Notes:
- agent: File mention chips in the chat input are now clickable and will
open the referenced files in the editor.
Closes #46746
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
## **Description:**
**Context:**
This PR introduces an initial CSV preview feature for Zed, building upon
two previously merged infrastructure PRs:
- [#46341](https://github.com/zed-industries/zed/pull/46341) - Data
table dynamic column support (removed const generics)
- [#46190](https://github.com/zed-industries/zed/pull/46190) - Variable
row height mode for data tables
This implementation is based on the [original draft PR
#44344](https://github.com/zed-industries/zed/pull/44344), which has
been carefully decomposed into smaller, reviewable pieces.
---
#### **Features Included:**
**Core Infrastructure:**
- Live CSV parsing with smart debouncing (200ms cooldown)
- Performance monitoring with built-in timing metrics (not displayed in
UI yet)
- Automatic file change detection and re-parsing
- Support for quoted fields, multiline cells, and escaped characters
**Table Display:**
- Variable row height rendering with fallback to uniform mode
(switchable via settings)
- Draggable column resizing (reusing existing data table infrastructure)
- Row identifiers supporting both source line numbers and sequential row
numbers
- Configurable font rendering (UI font vs monospace)
- Tooltips showing full cell content on hover
**Interactive Features:**
- Column sorting (ascending/descending) with visual indicators
**Settings Panel:**
- Toggle between variable/uniform row rendering
- Font type selection (UI/monospace)
- Row identifier type configuration
- Debug information display
- Multiline cell rendering options
---
#### **Features Intentionally Removed for This PR:**
To reduce complexity and review scope, the following features were
temporarily reverted and will be reintroduced in subsequent PRs:
- ❌ Settings pannel with performance metrics overlay
- ❌ Cell selection (single, multiple, and range selections)
- ❌ Keyboard navigation with arrow keys and selection extension
- ❌ Copy functionality supporting CSV, TSV, and Markdown table formats
- ❌ Inline cell editing with file persistence
- ❌ Viewport following for large datasets
- ❌ Column filtering and search capabilities
These removals were done via "time-machine" commits that cleanly nuked
vertical slices of functionality from the complete implementation.
---
**Technical Implementation:**
The feature is organized into a dedicated `csv_preview` crate with the
following structure:
```
crates/csv_preview/
├── src/
│ ├── csv_preview.rs # Main view and coordination logic
│ ├── parser.rs # CSV parsing and editor integration
│ ├── settings.rs # Configuration types and defaults
│ ├── table_data_engine.rs # Data transformation logic
│ ├── renderer/ # UI rendering modules
│ │ ├── preview_view.rs # Main render implementation
│ │ ├── render_table.rs # Table component assembly
│ │ ├── table_cell.rs # Individual cell rendering
│ │ ├── table_header.rs # Header with sorting controls
│ │ └── row_identifiers.rs # Line number column
│ └── types/ # Core data structures
│ ├── table_like_content.rs
│ ├── coordinates.rs # Display vs data coordinate systems
│ └── table_cell.rs
```
**Key architectural decisions:**
- **Dual coordinate system**: Separates data indices from display
indices to support sorting/filtering
- **Component reuse**: Leverages existing `data_table` infrastructure
from the keymap editor
---
**Integration:**
- Registers `csv::OpenPreview` action (currently without default
keybindings)
- Follows the same workspace integration pattern as `markdown_preview`
and `svg_preview`
- Automatically detects `.csv` file extensions
- Tab integration with appropriate icons and naming
---
**Code Structure Note:**
Some code structures, types, and documentation may appear redundant or
over-engineered in this initial implementation. This is intentional -
the feature was developed as a complete system and then decomposed by
functionality rather than being built incrementally. The "extra"
infrastructure supports features that were removed for this PR but will
be reintroduced in subsequent ones.
This approach was chosen over extensive refactoring because:
1. The complete feature took 200+ commits to develop with significant
rewrites
2. Clean extraction of vertical slices was more feasible than rebuilding
incrementally
3. The end state will utilize all these components, making current
"redundancy" temporary
I apologize for any inconvenience this may cause during review, but the
alternative would have required significant refactoring effort just to
make intermediate states "prettier," which seemed counterproductive.
---
**Future Work:**
This lays the groundwork for upcoming PRs that will reintroduce the
removed features:
- Cell selection and keyboard navigation
- Copy functionality with multiple output formats
- Inline editing capabilities with undo/redo
- Column filtering and search
- TSV and other delimiter support
- Improved horizontal scrolling behavior
- Settings persistence
**Testing:**
Includes test fixtures demonstrating multiline cell handling, various
column counts, and edge cases.
---
**Release Notes:**
- N/A This is feature flagged
---------
Co-authored-by: Anthony Eid <anthony@zed.dev>
Closes#50569
Before you mark this PR as ready for review, make sure that you have:
- [x] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
- [x] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
Previously, OpenRouter responses containing only usage data (without any
choices) would cause an error. Now the mapper properly emits usage
updates
for these events without failing.
Release Notes:
- Fixed an error when OpenRouter returns a usage-only event with empty
choices.
Moves loading of the history connection to once we have a new
connection, not on every thread view.
Release Notes:
- N/A
---------
Signed-off-by: rabi <ramishra@redhat.com>
Co-authored-by: Ben Brandt <benjamin.j.brandt@gmail.com>
When a completion request fails with a retryable error (e.g. a 500 from
the upstream provider), the retry loop waits on a timer before trying
again. This timer did not race with the cancellation signal, so if the
user switched models and submitted a new message during the retry delay,
the old turn would continue retrying with the stale model for up to 15
seconds — making requests to the wrong provider and corrupting the
thread's message list with spurious Resume entries.
Now the retry delay races with the cancellation receiver, so the old
turn exits immediately when cancelled.
Release Notes:
- Fixed cancelled turns in a conversation that failed (e.g. 500 from the
LLM provider) bein retried even after cancellation
Removes unused `edit_agent_output` from `EditFileTool`. This makes it
easier to maintain compatibility between the `EditFileTool` and
`StreamingEditFileTool`.
Release Notes:
- N/A
A vector was being instantiated when the callsite only required an
iterable.
Another part of the code was performing multiple `contains()` lookups on
a vector, and now it does it on a hashed set. This change has required
some extra modifications across the codebase, but the affected sites are
minimal and have been adjusted without major issues.
The changes include some `Hash` derived implementations, which were
proposed in the original `lsp-types` in [[1]], and maybe could be merged
into Zed's fork. I went ahead and used a newtype with a custom `Hash`
implementation that simply called on the structure's public members'
implementations of `Hash`.
The next change includes the removal of a check of the request
capabilities after having already checked the same thing in the call to
`to_lsp_params_or_response()` right before. The result of the `match`
expression should already have returned a `Task::ready()` if the above
mentioned function failed in performing the check that was later
repeated and now removed.
Finally, in the `edits_from_lsp()` method, stable sorting was being
performed when only unstable sorting would suffice. The method can only
sort with respect to the key data, and not the satellite data, as the
latter are the literal strings of the edit. It matters not which one of
a sequence of overlapping edits (with same ranges that thus resolve the
edits for equivalence) should come before the other.
[1]:
https://github.com/gluon-lang/lsp-types/pull/295/changes#diff-b1a35a68f14e696205874893c07fd24fdb88882b47c23cc0e0c80a30c7d53759R540
- [ ] Tests or screenshots needed?
- [ ] Code Reviewed
- [ ] Manual QA
Release Notes:
- Removed a vector allocation where the callsite only required an
iterable.
- Improved multiple lookup operations when deserializing LSP edit
operations.
- Removed a double-check of capabilities after requesting and thus
determining LSP capabilities.
- Replaced stable sorting with unstable sorting of edits returned by the
LSP.
Store the user's unsent message editor text in DbThread so it survives
quitting and reloading Zed. The draft flows through Thread → AcpThread →
AcpThreadView on load, and back via a debounced observer on the message
editor for saves.
Currently works for native Zed agents only; external ACP agents will
pick this up once general ACP history persistence lands.
## Changes
- **`DbThread`** / **`Thread`**: New `draft_prompt: Option<String>`
field, included in `to_db()`/`from_db()`
- **`AcpThread`**: Bridge field with getter/setter, populated during
`register_session()`
- **`NativeAgent::save_thread()`**: Copies draft from `AcpThread` →
`Thread` before persisting
- **`AcpThreadView`**: Restores draft into `MessageEditor` on load;
syncs editor text → `AcpThread` via observer; debounced (500ms) Thread
notify triggers DB save
Co-authored-by: Anthony Eid <anthony@zed.dev>
Co-authored-by: Mikayla Maki <mikayla@zed.dev>
When pinned tabs are present, activating an unpinned tab passed the
absolute tab index to the scroll handle, which only contains unpinned
tabs. This caused the scroll-into-view to silently fail.
Subtract `pinned_tab_count` from the index so it maps to the correct
child in the unpinned tabs scroll container.
Release Notes:
- Fixed tab bar not reliably scrolling to the active tab when pinned
tabs are present.
Closes#48812
## Summary
`editor::SplitSelectionIntoLines` currently triggers autoscroll and can
jump to the end of a long file.
This PR makes the action explicitly no-scroll.
## What changed
- Disabled autoscroll in the `unfold_ranges` call inside
`split_selection_into_lines`.
- Switched selection update to `SelectionEffects::no_scroll()`.
## Testing
- Added `test_split_selection_into_lines_does_not_scroll`.
### Screenshots:
- Before: see issue video
- After:
<img width="2486" height="1299" alt="image"
src="https://github.com/user-attachments/assets/8408a2f1-7dd7-4984-aa89-927d5a588322"
/>
Before you mark this PR as ready for review, make sure that you have:
- [x] Added a solid test coverage and/or screenshots from doing manual
testing
- [x] Done a self-review taking into account security and performance
aspects
- [x] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
Release Notes:
- Stop scrolling on `editor::SplitSelectionIntoLines` called in the long
files
