cc @SomeoneToIgnore
## Summary
Follow-up to [this
discussion](https://github.com/zed-industries/zed/discussions/55352#discussioncomment-16919854).
This extracts the edit prediction and clipboard code from `editor.rs`
into `edit_prediction.rs` and `clipboard.rs`.
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
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
Closes#53604.
When `log.showSignature = true`, git prepends signature verification
lines to stdout before the `--format` output. The null-separated parsers
in `crates/git/src/repository.rs` weren't expecting that, so SHAs ended
up corrupted — the git graph detail panel showed "0 changed files" for
every commit, and file history was similarly broken.
Setting `-c log.showSignature=false` in `build_command` is the same
trick we already use for `core.fsmonitor`. It only affects
log/show/whatchanged, so other commands aren't touched.
Verified locally with an SSH-signed repo: before this change the detail
panel said "0 changed files"; after, the modified files show up
correctly.
Release Notes:
- git_graph: Fix breakage that occurs when `log.showSignature` is
enabled
---------
Co-authored-by: Anthony Eid <anthony@zed.dev>
Seems to have been introduced in
1fbe1e3512.
These settings actually mean the opposite:
> Whether to use the system provided dialogs for Open and Save As. When
set to false, Zed will use the built-in keyboard-first pickers.
> Enables the simple file dialog for opening and saving files and
folders. The simple file dialog replaces the system file dialog when
enabled.
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
~~Closes #ISSUE~~
Release Notes:
- Fixed import of VSCode setting `files.simpleDialog.enable` being
inverted
## Summary
- Changes the outer `MarkdownPreviewView` div from `.size_full()` to
`.w_full().flex_1().min_h_0()`.
- Before: preview's height stayed at 100% of the pane regardless of
toolbar size, so deploying the buffer search bar pushed the preview's
bottom 45px past the pane edge — clipped by the pane's
`overflow_hidden`.
- After: preview takes only the leftover main-axis space and shrinks as
the toolbar grows. Bottom edge stays at the pane bottom.
Verified empirically: `bounds.size.height` round-trips `952px ↔ 907px`
with the toolbar at `67 ↔ 112` (Δ = 45px = the search bar's height).
Companion to #55008 (anchor compensation). Both PRs come out of
investigating #54987 — they fix two independent bugs that surface
together when toggling the search bar.
## Test plan
- [x] Compiles with `cargo clippy -p markdown_preview -- --deny
warnings`.
- [x] Manual: `Cmd-F` in the markdown preview no longer clips bottom
content past the pane edge.
- [x] Manual: `Esc` restores the preview to full available height.
Release Notes:
- Fixed markdown preview content being clipped past the pane edge when
the buffer search bar is open
Co-authored-by: Agus Zubiaga <agus@zed.dev>
<img width="541" height="245" alt="SCR-20260515-phfs"
src="https://github.com/user-attachments/assets/30413912-5c83-49c1-8874-4f987af222d3"
/>
Might be nice to in the future, have this open a modal to allow
inputting, similar to configuring MCP servers, but this is nice quick
fix for now.
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:
- N/A
Now that we have per-channel docs, it'll be nice to have a method to
link to "this" channel's docs when we ship features that link out to
them.
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
Replaces the `@rule` autocomplete entry in the agent chat input with
`@skill`, since Zed Rules are being deprecated in favor of Skills.
Behavior matches the old `@rule` flow: selecting a skill inserts a Skill
mention crease that loads the skill file as context, and the entry only
appears when there are available skills.
The underlying `MentionUri::Rule` parsing/handling is left in place so
previously saved rule mentions still load correctly — only the
autocomplete surface area was migrated.
Release Notes:
- Replaced the `@rule` autocomplete entry with `@skill` in the agent
chat input.
## Summary
- Document the new ChatGPT Subscription provider from #53166
- Add to the LLM Providers page alphabetically between Anthropic and
DeepSeek
- Cover sign-in flow, available models (GPT-5.5, GPT-5.3 Codex), and
sign-out
## Test plan
- [ ] Verify section renders correctly in docs preview
- [ ] Confirm links and anchors work
Release Notes:
- N/A
Hides the legacy Rules surfaces when the `skills` feature flag is
enabled, since Rules are surfaced through the Skills UI in that case:
- The **Rules** entry in the agent panel's ellipsis (triple-dot)
dropdown menu is hidden.
- The `assistant: open rules library` command palette action is hidden.
The `@rules` autocomplete in the chat input is intentionally left alone
for now.
### Implementation notes
`cx.has_flag::<SkillsFeatureFlag>()` reads from a global feature-flag
store that is populated asynchronously from the server:
- **Command palette**: `update_command_palette_filter` is already re-run
via `cx.on_flags_ready` once flags arrive, so the filter is reapplied
with the correct value. No new wiring needed.
- **Dropdown menu**: the popover's menu closure runs on every open, so
the latest flag value is reflected at click time.
Release Notes:
- N/A
This PR removes some more unused fields from the database user model:
- `github_user_created_at`
- `email_address`
- `name`
- `created_at`
These fields were not being used anywhere, and are nullable/defaulted in
the database in tests.
Release Notes:
- N/A
Sometimes the action log would not auto-accept deleted hunks, and I
finally found a repro for this, here's an explanation of what went wrong
```rust
// Before
use crate::{Alpha, Beta};
fn keep() {
work();
}
fn remove() {
work();
}
fn after() {
work();
}
```
```rust
// After commit
use crate::{Alpha};
fn keep() {
work();
}
fn after() {
work();
}
```
The action log may track the deletion as:
```diff
fn keep() {
work();
}
-fn remove() {
- work();
-}
-
fn after() {
work();
}
```
But the commit diff may choose different boundaries because nearby lines
repeat:
```diff
fn keep() {
work();
-}
-
-fn remove() {
- work();
}
fn after() {
work();
}
```
Both diffs produce the same final file, but their row ranges differ. The
previous logic only accepted committed edits when those row ranges
matched exactly, so already-committed edits could remain marked as
unaccepted.
Now we have a fast path for checking if the base_text matches exactly,
which works fine in this case.
Release Notes:
- Fixed an issue where agent edits would sometimes not get auto-accepted
when they were commited
When the Agent Panel had the Settings overlay open on top of the
ephemeral new-draft thread, pressing **Cmd-N** (or clicking *Start New
Thread* in Settings) silently did nothing — the action ran, but the
Settings overlay stayed painted on top.
The bug was in `AgentPanel::activate_draft`. Every other `activate_*`
path goes through `set_base_view`, which clears the overlay. But
`activate_draft` has an early return for the case where the base view
already holds the requested draft, and that early return only refocused
— it never touched `overlay_view`. So when the user was already on the
draft and the Settings overlay was on top, "new thread" had nothing
visible to do.
Fix: in that early-return branch, if an overlay is currently open, clear
it (which also handles focusing and emits `ActiveViewChanged`).
Added a regression test that puts the panel on the draft, sets the
Settings overlay, dispatches `NewThread`, and asserts the overlay is
gone. The test directly toggles `set_overlay(OverlayView::Configuration,
…)` rather than calling `open_configuration`, because the latter builds
provider configuration views and panics on the fake LM provider's
`unimplemented!()` `configuration_view` — and the bug being exercised
lives entirely in the overlay/base-view bookkeeping.
Closes AI-237
Release Notes:
- Fixed pressing "New Thread" (or Cmd-N) doing nothing when the Agent
Panel's Settings view was open
wires user configured `FontFallbacks` into the cosmic text path. the
chain is resolved at font load time and stored on each `LoadedFont`.
`layout_line` splits each `FontRun` into spans by codepoint coverage and
emits one `Attrs` per slot so cosmic text shapes each span with the
correct face. inheriting codepoints (marks, zwj, zwnj, variation
selectors) stick to the current span so emoji zwj sequences and
combining marks are not torn across faces.
Closes#17254
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] Unsafe blocks (if any) have justifying comments
- [x] Performance impact has been considered and is acceptable
- [x] Tests cover the new/changed behavior
- [] The content is consistent with the [UI/UX
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
Release Notes:
- added support for `buffer_font_fallbacks` 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
- [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
Closes #ISSUE
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
Closes #ISSUE
Release Notes:
- N/A or Added/Fixed/Improved ...
I've been working on a GPUI application which has a button for copying
text. After starting the app in Wayland in Linux I noticed that when I
click the button to copy text it does not work. After interacting with
other buttons and/or copying text via keyboard shortcuts then my copy
button works.
I thought I was handling something wrong, but then I noticed that Zed
also exhibits this behavior: upon starting Zed if you highlight text and
right-click and select "Copy" then nothing happens. After you interact
with other UI elements and/or keyboard shortcuts then copy buttons seem
to work fine.
This paper cut has been annoying me and with the help of Claude Code I
arrived at this small fix. This solves the problem and this appears to
be the way the serial should be handled for clipboard actions via button
clicks in Wayland, as far as I've been able to learn. Here are a couple
related Wayland documentation pages that I double checked:
1.
https://wayland.app/protocols/wayland#wl_data_device:request:set_selection
2.
https://wayland.freedesktop.org/docs/html/ch04.html#sect-Protocol-data-sharing-devices
- [x] Added ~a solid test coverage and/or~ screenshots from doing manual
testing
- Here's a screen recording of the behavior without this fix:
[wayland-clipboard-problem.webm](https://github.com/user-attachments/assets/6c7d3b3b-56fe-4083-a011-4906ee9bfbec)
- Here's a screen recording of the behavior with this fix:
[wayland-clipboard-fix.webm](https://github.com/user-attachments/assets/167ee731-118a-4f67-8489-6cfaca4389ce)
- Here's a screen recording of the problem in Zed itself:
[wayland-clipboard-problem-zed.webm](https://github.com/user-attachments/assets/e41a40e1-54aa-4b3f-a2c6-f3e90d06d50f)
- [x] Done a self-review taking into account security and performance
aspects
- None that I could find or think of
- [x] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
- I don't believe this is applicable...
Release Notes:
- Improved first mouse-driven text copy action on Wayland
Settings search treated multi-word queries as OR, so “tab size” matched
any document whose words contained a prefix of “tab” or “size” --
flooding results with unrelated entries like “Show Tab Bar” and “Maximum
Tabs”.
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)
- [ ] Tests cover the new/changed behavior
- [ ] Performance impact has been considered and is acceptable
Closes#56244.
Release Notes:
- Fixed settings UI search returning unrelated results for multi-word
queries
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
The idle guard prevented the keybinding from working while the agent was
generating, even though the profile selector button in the UI remained
clickable. Changing the profile affects future messages, not the
in-flight response, so there is no reason to block it.
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
Closes#56704
Release Notes:
- Fixed the `agent: toggle profile selector` keybinding being ignored
while the agent was generating a response.
---------
Co-authored-by: Danilo Leal <67129314+danilo-leal@users.noreply.github.com>
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
Closes: #56857 Title generation failing on agent threads with OpenAI
models from subscription
Currently title generation requests are made without any system prompt
which causes instructions to be not set, and the request fails. This
change makes sure there is always instructions set (api does not seem to
care if its empty or not)
Release Notes:
- N/A
Follow-up to https://github.com/zed-industries/zed/pull/56743 where we
only show the focused styles in history tab items if you're navigating
it with the keyboard. We were previously flashing the focused border
styles on click and they're not necessary if you're pointer-based.
Release Notes:
- 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 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: #56844
Release Notes:
- Added support for specifying effort level when using OpenAI models via
ChatGPT subscription
Follow up to https://github.com/zed-industries/zed/pull/56689. This PR
fixes an error where any instance of a slash character in the message
editor was going through the slash command validation. The fix makes the
validation get triggered only if the slash character is the very first
thing typed in the message. This follow the approach Slack, Discord, and
other apps that contain slash commands use. Also tweaked the messaging a
little bit so it's crystal clear what's happening.
Release Notes:
- N/A
- Subscribe to buffer LanguageChanged events so language detection that
completes later (e.g. in remote projects) triggers a kernelspec refresh
and ensures the REPL UI is populated.
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
Closes#54388
Release Notes:
- Python REPL UI stays responsive on remote connection
---------
Co-authored-by: MrSubidubi <finn@zed.dev>
The fix comes based on that if the button is already wrapped in the
popover menu, it doesn't need to have a click handler for the menu to
open!
Release Notes:
- FIxed bug where the activity indicator in the status bar wouldn't open
a menu with menu items to cancel running actions.
Fixes editor hit testing when the editor is clipped, most easily
noticeable in a partially off-screen markdown block in the agent panel.
`PositionMap::point_for_position` was indexing into `self.line_layouts`
relative to the editor's scroll position (which in this example doesn't
apply because these editors don't scroll), rather than relative to
`self.visible_row_range.start`. This caused the reported column to not
always be computed using the correct row, which especially impacted rows
closer to the bottom of the editor. Notice how closer to the bottom,
`point_for_position` reports the correct row, but reports the column as
0:
https://github.com/user-attachments/assets/2a5b2c8b-2994-4983-a33b-ede209ba00e6
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:
- Fixes selection not always matching the mouse position in a markdown
block in the agent panel
## Summary
Adds column pinning (freeze) capability to data tables, allowing the
first N columns to remain visible while scrolling horizontally through
the rest of the table content.
Common spreadsheet/data table UX pattern. When viewing wide tables with
many columns, users need to see identifying information (row labels,
IDs) while scrolling right to explore data.
## Demo
Idle state:
<img width="559" height="179" alt="image"
src="https://github.com/user-attachments/assets/b3f89221-8aa9-4e8a-9a39-6f06cc8a4eea"
/>
Scrolled horizontally (name column dissapeared, line numbers column
stayed pinned):
<img width="522" height="174" alt="image"
src="https://github.com/user-attachments/assets/c6695a00-5e40-49b8-81d7-0b30e26bb7bb"
/>
## Implementation
- New `pin_cols(n: usize)` builder method on `Table` to specify how many
columns to pin
- Pinned columns render in a fixed section that doesn't scroll
horizontally
- Scrollable columns render separately with independent scroll state
- Horizontal scroll offset adjustments for proper column resize handle
positioning with pinned sections
- Pinned section stays at viewport left edge while scrollable section
scrolls independently
- Supports 0 < pinned_cols < total_cols (partial pinning)
- Applied to CSV preview for better UX with wide datasets
## Context
Part of CSV preview feature series, following PR #53496 (settings UI).
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
- [ ] ~~Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)~~
no ui changes besides pinning itself. UI improvements out of scope of
this PR
Release Notes:
- Improved CSV preview with column pinning to keep identifiers visible
while scrolling
---------
Co-authored-by: Smit Barmase <heysmitbarmase@gmail.com>
This updates the nightly workflow to run hourly as opposed to once
daily.
Notably, it also
- changes the poll duration for the auto updater - we now poll 4 times
an hour on the Nightly channel and continue with the old rate for all
other channels
- moves the pre-checks to the linux runners, since we have more
resources available there
- only ever allow one concurrent nightly run
Release Notes:
- N/A
https://github.com/zed-industries/zed/actions/runs/25881329418 complains
about the deprecation of Node.js version 20, which will soon be
force-bumped to 24 in the runner contexts.
Thus, this PR bumps all the jobs mentioned there to versions that run on
Node.js 24 by default. I checked all the release note section for the
three actions and this should not change any behaviour according to
their release notes and get us ready for the deprecation.
Release Notes:
- N/A
Closes AI-230
This PR makes skills, added as /-mentions, be rendered in the agent
panel as creases, like anything you'd @-mention. Naturally, clicking on
the crease button opens the corresponding skill file in a buffer.
It turned out to be quite a bit of plumbing to make this work,
particularly as I am also introducing an interface to display dividers
and headers in the completion menu. This was relevant to me to add
because it sets a good foundation to convert many agent panel-related
actions as slash commands.
Release Notes:
- N/A
---------
Co-authored-by: MartinYe1234 <52641447+MartinYe1234@users.noreply.github.com>
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
Closes https://github.com/zed-industries/zed/issues/55746
Release Notes:
- Fixed an issue where file system events or language server events that
resulted in Zed unwatching many paths would result in high CPU usage
Adds a new language model provider that lets users authenticate with
their ChatGPT Plus/Pro subscription and use OpenAI models
(codex-mini-latest, o4-mini, o3) directly in the Zed agent — without
needing a separate API key.
## How it works
1. **OAuth 2.0 + PKCE sign-in**: Uses OpenAI's official Codex CLI client
ID to run an authorization code flow. A local HTTP server on
`127.0.0.1:1455` captures the callback, exchanges the code for tokens,
and stores them in the system keychain.
2. **Token refresh**: Access tokens are automatically refreshed when
they're within 5 minutes of expiry, using the stored refresh token.
3. **Responses API**: Requests go to
`https://chatgpt.com/backend-api/codex/responses` using the existing
`open_ai::responses` client (Responses API format, not Chat Completions
which was deprecated for this endpoint in Feb 2026).
4. **Required headers**: `originator: zed`, `OpenAI-Beta:
responses=experimental`, `ChatGPT-Account-Id` (extracted from JWT),
`store: false` in the body.
## Files changed
- `crates/open_ai/src/responses.rs`: Add `store: Option<bool>` field to
`Request`; add `extra_headers` param to `stream_response` for
per-provider header injection
- `crates/language_models/src/provider/openai_subscribed.rs`: New
provider (sign-in UI, OAuth flow, token storage/refresh, model list)
- `crates/language_models/src/provider/open_ai.rs`,
`open_ai_compatible.rs`, `opencode.rs`: Pass `vec![]` for new
`extra_headers` param
- `crates/language_models/src/language_models.rs`: Register the new
provider
- `crates/language_models/Cargo.toml`: Add `rand` and `sha2` deps for
PKCE
## Open questions / known gaps
- [ ] **Terms of service**: Usage appears to be within OpenAI's ToS
(interactive use via their official CLI client ID), but needs legal
sign-off before shipping
- [ ] **Redirect URI**: Currently `http://localhost:1455/auth/callback`
— may need to match exactly what OpenAI's Codex CLI uses
- [ ] **UI polish**: The sign-in card is functional but minimal; needs
design review
- [ ] **Error messages**: OAuth error responses from the callback URL
aren't surfaced to the user yet
- [ ] **`o3` availability**: o3 may require a higher subscription tier;
consider gating it
## Testing
Sign-in flow was designed to match the Copilot Chat provider pattern.
Manual testing against the live OAuth endpoint is needed.
Release Notes:
- Added ChatGPT subscription provider, allowing users to use their
ChatGPT Plus/Pro subscription with the Zed agent
---------
Co-authored-by: Zed Zippy <234243425+zed-zippy[bot]@users.noreply.github.com>
Co-authored-by: Richard Feldman <richard@zed.dev>
Co-authored-by: Richard Feldman <oss@rtfeldman.com>
Co-authored-by: Agus Zubiaga <agus@zed.dev>
Adds a `markdown_preview_code_font_family` setting that overrides the
font used in the markdown preview for code — inline `code` spans and
fenced code blocks.
This is the counterpart to `markdown_preview_font_family` (#54003),
which already does this for body text. Together they let you choose a
typographically matched body + code font pair for the preview without
forcing the code font onto editor buffers, where it may not be a good
coding font:
```json
{
"markdown_preview_font_family": "Noto Serif",
"markdown_preview_code_font_family": "Noto Sans Mono"
}
```
Behavior mirrors `markdown_preview_font_family`:
- Scoped to the markdown preview only (`MarkdownFont::Preview`). The
agent panel, notifications, hover popovers, and REPL output are
unaffected — they keep using the buffer font for code.
- Falls back to the buffer font family when unset, so existing previews
are unchanged.
- Overrides the font family only; fallbacks and features still come from
the buffer font.
Before (uses buffer font, here Iosevka):
<img width="509" height="368" alt="Screenshot 2026-05-14 at 1 39 51 PM"
src="https://github.com/user-attachments/assets/6b7e49b2-fc6e-4db1-9679-392b3447f411"
/>
After (uses specified font, here Noto Sans Mono):
<img width="508" height="368" alt="Screenshot 2026-05-14 at 1 40 51 PM"
src="https://github.com/user-attachments/assets/f911c99b-08f8-4336-83eb-54b555f11c54"
/>
Release Notes:
- Added `markdown_preview_code_font_family` to override the code font in
the markdown preview
It's possible to configure `sort_mode` for the project panel like that:
```json
"project_panel": {
"sort_mode":"directories_first"
}
```
However, `workspace: open files` doesn't respect this setting and always
shows files and directories mixed. This change fixes the issue.
Release Notes:
- Improved sorting in `workspace: open files`
---------
Co-authored-by: Christopher Biscardi <chris@christopherbiscardi.com>
Added a hint for `use_podman` setting.
Self-Review Checklist:
- [x] I've reviewed my own diff for quality, security, and reliability
- [x] 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)
- [ ] Tests cover the new/changed behavior
- [ ] Performance impact has been considered and is acceptable
Closes (issue to be created, but it's just about an outdated doc)
Since we download the grammars into the `./grammars` directory, the
formatting check can fail on unformatted query files that are present in
the grammars directory, as observed in
https://github.com/zed-extensions/log/pull/18
Thus, this PR changes the check to be more specific to just the queries
in the `languages` directory.
Release Notes:
- 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 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
fixvuejs/language-tools#5956
Align with
3fb3329dfb/extensions/vscode/src/extension.ts (L223).
Release Notes:
- Always pass null when body is non-existent in Vue language server
request
---------
Co-authored-by: Smit Barmase <heysmitbarmase@gmail.com>
We originally added support for these keys to allow people to customize
the pill. As I noted at the time, we'd like the default to be more
visually minimal.
Release Notes:
- Reverted the colored vim mode indicator pill in the One Light and One
Dark themes (#56662)
Adds a one-time, idempotent startup migration that moves every user Rule
out of the `PromptStore` LMDB database into the new Skills + AGENTS.md
world, in a single pass:
- **Non-Default Rules → global Skills.** Each one becomes
`~/.agents/skills/<slug>/SKILL.md` with `disable-model-invocation:
true`, preserving the original behavior that non-Default Rules were only
ever invoked when the user named them. They're now invokable via
`/skill-name` (and still `@`-mentions).
- **Default Rules → global AGENTS.md.** Each one is appended to
`paths::agents_file()` (e.g. `~/.config/zed/AGENTS.md` on macOS/Linux,
`%APPDATA%\Zed\AGENTS.md` on Windows) under an `## H2` heading
containing the rule's title. Default Rules used to be auto-included in
every conversation; the global AGENTS.md is loaded into the system
prompt of every conversation, so the behavior is preserved.
- **Customized built-in prompts → global AGENTS.md** (currently just the
commit-message prompt). If the user has edited a built-in away from
Zed's shipped `default_content()`, the edited body is appended at the
top of the AGENTS.md block. Uncustomized built-ins (still on the shipped
default) are skipped so we don't pollute AGENTS.md with text the user
never wrote.
The migration is gated on the `skills` feature flag — users without the
flag never have their Rules touched in any way. A single global KVP flag
(`rules_to_skills_migration_done`) short-circuits the migration on
subsequent launches, so it runs at most once per machine even across
release channels. A process-lifetime `AtomicBool` guard additionally
prevents racing duplicate spawns when the underlying `cx.on_flags_ready`
callback fires multiple times at startup.
Migration is intentionally non-destructive: rule rows in the LMDB
database stay in place. Users can still see and edit them through the
existing UI, and a downgrade to a Zed build without skills support won't
lose anything.
Slug generation (`agent_skills::slugify_skill_name`) lowercases ASCII
letters, turns spaces into dashes, and drops every other
non-alphanumeric character entirely — so `foo!bar` becomes `foobar`, not
`foo-bar`. `&` is special-cased to become `and` (so `rock&roll` →
`rock-and-roll`). Slug collisions and pre-existing skill directories are
handled by appending `-2`, `-3`, etc.
A title-bar onboarding banner ("Skills have replaced Rules") surfaces
for every user on the `skills` feature flag. Clicking it opens a small
`AlertModal`-based explainer that summarizes the two destinations and
points users at the new `/skill-name` slash command (and notes that
`@`-mentions still work).
Closes AI-227
Closes AI-232
Release Notes:
- N/A
## Context
<!-- What does this PR do, and why? How is it expected to impact users?
Not just what changed, but what motivated it and why this approach.
Link to Linear issue (e.g., ENG-123) or GitHub issue (e.g., Closes#456)
if one exists — helps with traceability. -->
Closes https://github.com/zed-industries/zed/issues/52242
This PR reverts a change that highlights shell script arguments as
variables rather than strings. Because of how shell scripts work, all
parameters are strings by default and are regularly represented as such.
The original PR (https://github.com/zed-industries/zed/pull/48067) aimed
to improve highlighting by capturing all arguments as
`variable.parameter`s, but this doesn't align with the idiomatic usage
of shell scripts.
| Current Behavior | Expected Behavior |
|---|---|
| <img width="242" height="84" alt="Image"
src="https://github.com/user-attachments/assets/50c87198-e4be-4627-a742-88060b2b9d14"
/> | <img width="227" height="83" alt="Image"
src="https://github.com/user-attachments/assets/f08b1c62-9c7a-460e-b306-4ac6b9e5a897"
/> |
| Text arguments are highlighted as variable | Text arguments are
highlighted as "string" |
## Self-Review Checklist
<!-- Check before requesting review: -->
- [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
- [ ] Performance impact has been considered and is acceptable
Release Notes:
- N/A or Added/Fixed/Improved ...
Resolves https://github.com/zed-industries/zed/issues/56478
This diff fixes Bash syntax highlighting for conditional and arithmetic
expressions. The Bash parser already recognizes tokens like `[[`, `]]`,
`((`, `))`, `$((`, `!=`, and `||`, but Zed’s Bash highlight query was
not capturing several of them.
With this change, the Bash highlight query styles those operators and
delimiters consistently with existing shell operators, brackets, and
special punctuation.
| Before | After |
| --- | --- |
| <img width="1667" height="1088" alt="ayu-old"
src="https://github.com/user-attachments/assets/0a89bd71-0a13-4ebe-8bc9-abc27ab42ed8"
/> | <img width="1667" height="1088" alt="ayu-new"
src="https://github.com/user-attachments/assets/afa5e8ab-f94a-4f5e-8fe8-770bac07559a"
/> |
Release Notes:
- Fixed bash syntax highlighting for conditional expressions, arithmetic
expressions, and related operators.
The `draw()` method was calling `.pop()` on `next_frame.input_handlers`
to extract the active input handler for the platform window. This
reduced the Vec length, making cached `paint_range` indices stale. On
the next frame, when a cached view called `reuse_paint()`, it would
index into `rendered_frame.input_handlers` with out-of-bounds indices,
causing a panic: `range start index 1 out of range for slice of length
0`.
**Fix:** Use `.last_mut().and_then(|h| h.take())` instead of `.pop()` to
extract the handler without changing the Vec length. The slot becomes
`None`, which is already handled by `reuse_paint`'s `.take()` logic.
Closes#50456
### Testing
- All 83 existing GPUI unit tests pass
- Manually verified with
[longbridge/gpui-component](https://github.com/longbridge/gpui-component)
`cargo run --example dock`:
- Switch to Input panel → type text → no crash ✅
- Before fix: crash immediately on first keystroke
---
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
- [ ] Aligned any UI changes with the [UI
checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist)
*(N/A — no UI changes)*
Release Notes:
- Fixed a crash in GPUI when typing into an Input widget inside a cached
view ([#50456](https://github.com/zed-industries/zed/issues/50456))
Co-authored-by: Mikayla Maki <mikayla@zed.dev>
