This reverts commit 305b27738ef03438540c07301aa8f0f51aa79d92.

Drop the `count_tokens` API and related implementations across
providers, and remove the unused `tiktoken-rs` dependency.

I was going to update the dependency becuase they finally released a fix
we needed. But then I realized we only used this api in one place, the
Rules library. And for most models it would have been wildly incorrect
becuase we use tiktoken, i.e. OpenAI tokenizers, for almost every model,
which is going to give incorrect results.

Given that, I just removed these because the difference in how we get
these has caused plenty of confusion in the past.

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

Reverts zed-industries/zed#54475

This causes inline code blocks to look strange in some themes

Release Notes:

- Fixed Helix vgl now correctly selects to end of line without
including the newline character in the selection.

---------

Co-authored-by: Jakub Konka <kubkon@jakubkonka.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 #ISSUE

Release Notes:

- N/A or Added/Fixed/Improved ...

The "Command menonics" section in Zed's vim mode documentation made it
seem that the mentioned aliases were available, by default, which isn't
the case. These changes update the documentation to make it clearer that
this is meant to be set on the `command_aliases` setting and provides
and example configuration that users can easily copy and paste into
their settings file.

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 #52894

Release Notes:

- N/A

---------

Co-authored-by: dino <dinojoaocosta@gmail.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 #ISSUE

Release Notes:

- N/A or Added/Fixed/Improved ...

`SendKeystroke` seems to work for all ctrl keybinds. Prefer it over
`SendText` for readability.

showkey reports the same key sequences if I apply these keybinds to my
user 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
- [ ] 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

Co-authored-by: Jakub Konka <kubkon@jakubkonka.com>

showkey in Ghostty, iTerm2, and Terminal.app reports ctrl-delete as
`<ESC>[3;5~`.

I tested this keybind on macOS, where fish_key_reader correctly
interprets it as ctrl-delete.

Closes #51725

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:

- Fixed ctrl-delete keybind in the terminal

---------

Co-authored-by: Jakub Konka <kubkon@jakubkonka.com>

Release Notes:

- Added interleaved_reasoning option to openai compatible models

---

This PR adds the interleaved_reasoning option for OpenAI-compatible
models, addressing the issue described in
https://github.com/ggml-org/llama.cpp/issues/20837.

In my testing, enabling interleaved_reasoning not only resolved the
tool-calling issues encountered by Qwen3.5 models in llama.cpp, but also
appeared to improve the model's coding capabilities. I have also
verified the outgoing requests using a proxy to ensure the parameter is
being sent correctly.It is also likely that this change will benefit
other models and providers as well.

Note: While I used AI to assist with the implementation, I have reviewed
and tested the changes. As I am relatively new to Rust and the Zed
codebase, I would appreciate any feedback or suggestions for
improvement. I am happy to make further adjustments if needed.

Thank you all for building such an amazing editor!

Co-authored-by: Oleksiy Syvokon <oleksiy@zed.dev>

Previously, sticky scroll headers were painted on top of the main
editor. This works, but only if the editor background is a solid color

Bad:
<img width="603" height="319" alt="image"
src="https://github.com/user-attachments/assets/bf438003-a416-46a6-85bc-dfa767be067e"
/>


Good:
<img width="603" height="319" alt="image"
src="https://github.com/user-attachments/assets/e644e54a-6667-41e1-86d0-16e659bea844"
/>

Release Notes:

- Fixed: Editor no longer renders behind sticky scroll headers while
using a transparent theme

I noticed that there are helper migration functions that apply settings
transformations to multiple layers (root, platform, release-channel, and
profile):

`migrate_settings`
`migrate_language_setting`

Both of these handle the old settings profile forms, so I've updated
them to handle either form, for usage in historical and future
migrations.

Migration `m_2026_04_10` actually introduced a wrapper function to
provide to `migrate_settings` to handle both forms, which is what clued
me into the fact that I missed these migration helpers when I wrote the
`m_2026_04_01` originally (I didn't know they existed).

Self-Review Checklist:

- [X] I've reviewed my own diff for quality, security, and reliability
- [N/A] Unsafe blocks (if any) have justifying comments
- [N/A] 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

---------

Co-authored-by: Smit Barmase <heysmitbarmase@gmail.com>

This fixes #53869.

Creating a new file from the project panel background menu failed when a
single local project was empty and its root was hidden. In that state
there are no visible entries to seed `expanded_dir_ids`, so the action
returned early before opening the filename editor.

This initializes that state lazily from the root entry when creating a
new item, and adds a regression test for the empty hidden-root path.

Release Notes:

- Fixed creating a new file from the project panel context menu in empty
local projects

Closes #48642.

- [ ] Tests or screenshots needed? - Tests written by AI. Not sure if
the test suite is setup correctly. I’ll come back later and see if the
tests are utter BS or not.
- [ ] Code Reviewed
- [x] Manual QA


Release Notes:

- Handle `menu::SecondaryConfirm` properly in the terminal

Co-authored-by: Bennet Bo Fenner <bennet@zed.dev>

Insert sessions before awaiting `session/load` so replayed
`session/update` notifications can find the thread.

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:

- acp: Fix for some replay events getting dropped when loading a
previous session.

## Summary

Inline code in rendered markdown previously used only the default
foreground color with a subtle background tint, which made it visually
indistinguishable from surrounding prose in some themes.

This change looks up the `text.literal` syntax style from the active
theme and applies its color to inline code, matching the treatment used
for inline code elsewhere in the editor (e.g. syntax highlighting in
source buffers). When the theme does not define `text.literal`, the
color falls back to the previous default.


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 inline code in rendered markdown to use the theme's
`text.literal` syntax color

Previously, passing an empty range to `range_to_buffer_range` would
return no ranges, which would trigger a fallback case in the git diff
that caused the entire multibuffer to be invalidated every frame.

This PR fixes `range_to_buffer_range`, so that it returns a sensible
range if the input range is empty, preventing the fallback behaviour and
edit expansion.

Self-Review Checklist:

- [ ] 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 #ISSUE

Release Notes:

- N/A or Added/Fixed/Improved ...

---------

Co-authored-by: Cole Miller <cole@zed.dev>
Co-authored-by: Anthony Eid <hello@anthonyeid.me>

Just a little refinement as I experienced the thread switcher modal
overflowing beyond the viewport if there were enough threads on it.
Additionally, took the opportunity to render a scrollbar inside of it,
too.

Release Notes:

- N/A

This PR removes the overlap between the sidebar scrollbar and the thread
item icon buttons. Now, they don't render on top of each other anymore
:)

Release Notes:

- Agent: Improved the scrollbar overlap with the thread item icon
buttons in the threads sidebar.

## Context

Heading sizes in the markdown preview were nearly identical to body
text, making `# H1`, `## H2`, and `### H3` visually indistinguishable.
The root cause was in `MarkdownStyle::themed()`: it set a
`heading_level_styles` override with font sizes of `rems(1.05–1.15)`,
which silently replaced the correct sizes applied by
`apply_heading_style` (via GPUI's `text_3xl`/`text_2xl`/`text_xl`
utilities `rems(1.875/1.5/1.25)`). Removing that override restores the
intended hierarchy. A `mt_4()` top margin was also added so consecutive
headings have visual breathing room.

Closes #54358

Video of manual test below :

[Screencast from 2026-04-21
02-22-12.webm](https://github.com/user-attachments/assets/8dd815f9-6f9b-4e88-bebb-28c79f019427)

## How to Review

`crates/markdown/src/markdown.rs` Three changes:
- (1) removed the `heading_level_styles` block from
`MarkdownStyle::themed()` that was overriding heading font sizes with
nearly-body-text values;
- (2) added `mt_4()` to the heading div in `push_markdown_heading` for
better vertical spacing;
- (3) added `test_heading_font_sizes_are_distinct` which renders H1–H3
and a paragraph then asserts that line heights strictly decrease from H1
down to body text.

## Self-Review Checklist

- [x] I've reviewed my own diff for quality, security, and reliability
- [ ] 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 heading sizes in the markdown preview to correctly reflect the
H1–H6 size hierarchy, matching standard markdown renderers

## Context

Closes #53741

Right-clicking on a link in any Markdown view showed no way to copy the
URL. The right-click handler already detected links for left-click
navigation, but the context menu was never extended to surface a
link-specific action.

Video of the manual test below :

[Screencast from 2026-04-13
00-29-49.webm](https://github.com/user-attachments/assets/fbde09ab-78da-4366-b1e0-e15e0d43442b)

## How to Review

- **`crates/markdown/src/markdown.rs`** — Added a `context_menu_link:
Option<SharedString>` field to `Markdown`. Added
`capture_for_context_menu(link)` (replaces the old
`capture_selection_for_context_menu`) which saves both selected text and
the hovered link together. Added a `context_menu_link()` accessor.
Updated the capture-phase right-click handler to detect the link under
the cursor via `rendered_text.link_for_source_index`. Added a
`event.button != MouseButton::Right` guard to the bubble-phase
`MouseDownEvent` handler to prevent selection logic from running on
right-click.

- **`crates/agent_ui/src/conversation_view/thread_view.rs`** — In
`render_message_context_menu`, after computing `has_selection`, also
reads `context_menu_link` from the same markdown chunks. Adds a "Copy
Link" entry with a separator at the top of the menu when a link URL is
present.

- **`crates/markdown_preview/src/markdown_preview_view.rs`** — Wraps the
markdown element in a `right_click_menu` with a "Copy Link" entry (when
a link is present).

Edit: There was a mention of a "Copy" and "Copy as Markdown" buttons.
After discussion, it was decided that I would re-add them fully fleshed
out in a separate PR

## Self-Review Checklist

- [x] I've reviewed my own diff for quality, security, and reliability
- [ ] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the UI/UX checklist
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable

Release Notes:

- Added "Copy Link" to the right-click context menu when clicking on a
  link in Markdown views (agent panel and Markdown preview)

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 #54419

Release Notes:

- Fixed Agent menu shortcut context handling so Alt-Shift-L consistently
resolves to Rules (and displays correctly) when the Toggle Agent Menu is
open, while preserving existing global/context-specific shortcuts such
as Language Servers and onboarding bindings outside Agent menu context.

---------

Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
Co-authored-by: Danilo Leal <67129314+danilo-leal@users.noreply.github.com>

The report summary previously assumed that every commit that had a
review would also have an associated PR - this was no longer the case
with Zippy PRs that are allowed without PR. This pull request fixes this
wrong assumption.

Release Notes:

- N/A

#### Context

When tree-sitter parses a file with broken syntax (e.g. a large partial
SQL `VALUES` clause, or any language where a large chunk becomes
invalid), it can produce a single `ERROR` node spanning thousands of
lines. On every render frame, Zed queries this tree for syntax
highlights via `SyntaxMapCaptures`. Previously, only `set_byte_range`
was applied to the query cursor - this limits which captures are
*returned*, but tree-sitter still had to *traverse* the entire ERROR
subtree to find them, causing O(file size) work per frame and making
scrolling/editing visibly laggy.

The fix applies `set_containing_byte_range` to the highlight query
cursor, mirroring what `SyntaxMapMatches` already does for indentation
and bracket queries. This tells tree-sitter to skip subtrees that extend
far beyond the visible window, reducing traversal to the visible range
only.

**Note:** This fix eliminates the main freeze/stall caused by full-tree
traversal. A small amount of lag may still occur on very large broken
files, as tree-sitter still needs to parse the error-recovery structure.
Further improvements would require deeper changes to tree-sitter's query
execution or incremental parsing.


#### Closes #52390

#### How to Review

Small change — focus on
[syntax_map.rs:1119-1123](crates/language/src/syntax_map.rs#L1119) (the
fix) and the `containing_byte_range_for_captures` helper below it.
Compare with the existing `SyntaxMapMatches::new` path (line ~1255)
which uses the same pattern.

#### 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

#### Video 
[Screencast from 2026-03-26
14-19-19.webm](https://github.com/user-attachments/assets/6628492a-f013-438a-836a-2740f6e2f266)


#### Note : Reopens previous work from closed PR #52475 (fork was
deleted)

Release Notes:

- Fixed laggy scrolling and editing in files with large broken syntax
regions (e.g. incomplete SQL `VALUES` clauses or large invalid blocks in
any language)

---------

Co-authored-by: Max Brunsfeld <maxbrunsfeld@gmail.com>

For docker-compose-based dev containers, build args live on the primary
compose service's `build.args`, not on `dev_container.build` (which is
`None` in the compose case). `expanded_dockerfile_content` only
consulted the latter, so `${…}` references in the service's Dockerfile —
e.g. `FROM ${BASE_IMAGE}` — were never substituted. Downstream callers
then invoked `docker inspect "${BASE_IMAGE}"` and `docker pull
"${BASE_IMAGE}"`, both of which fail and the dev container fails to
start.

A new unit test `test_expands_compose_service_args_in_dockerfile` mounts
a Dockerfile with `FROM ${BASE_IMAGE}` backed by a compose service whose
`build.args` define `BASE_IMAGE=test_image:latest`, and asserts both
`expanded_dockerfile_content` and `image_from_dockerfile` produce the
resolved reference.

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 dev container startup failing for docker-compose configs whose
service Dockerfile uses build-arg substitution in the FROM line (for
example, `FROM ${BASE_IMAGE}`).

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 #54424

Release Notes:

- N/A

Co-authored-by: Danilo Leal <67129314+danilo-leal@users.noreply.github.com>

This adds checks for the Zed Zippy version bumps. These are rather
strict, but since we should also be strict with the bumps themselves, I
consider this the better approach.

Note this intentionally only includes the version bumps here for now.

Release Notes:

- N/A

## Context

Fixes a Windows path handling bug in `crates/util/src/paths.rs` when
computing a relative path via `strip_prefix`. (Closes #51758 )

Previously, Windows prefix matching only handled drive-letter
case-insensitivity. It could still fail when the parent and child paths
referred to the same location but used different separator styles (`\`
vs `/`) or different casing in later path segments. That caused valid
Windows paths to return `None` instead of a relative path, which caused
the diff to break.

This change normalizes Windows paths for prefix comparison by
lowercasing any ASCII characters and converting backslashes to forward
slashes before checking the prefix.

Tests were added for mixed-case and mixed-separator Windows paths to
cover the bug and prevent further regressions.

## How to Review

1. Check the `strip_prefix` Windows branch to confirm the new
normalization logic only affects prefix comparison and still rejects
partial-segment matches.
2. Check the added tests for mixed separator and mixed casing cases on
Windows paths.

## 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:

- Fix blank git diff view on Windows when working in a subfolder

---------

Co-authored-by: John Tur <john-tur@outlook.com>

Closes #42731 

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)

<img width="1440" height="900" alt="Screenshot 2026-03-09 at 5 10 46 PM"
src="https://github.com/user-attachments/assets/bf124481-dc10-44e9-aaab-3e562d71e41e"
/>

Release Notes:

- Fixed Windows path handling in extension manifests to ensure
extensions upload correctly to remote environments like WSL.

---------

Co-authored-by: John Tur <john-tur@outlook.com>

The `ReviewBranchDiff`, `ResolveConflictsWithAgent`, and
`ResolveConflictedFilesWithAgent` actions are dispatched from the git UI
(`git_ui/src/project_diff.rs` and `git_ui/src/conflict_view.rs`), not
the agent panel. Their `source` field in the "Agent Thread Started"
telemetry event was set to `"agent_panel"` — this changes it to
`"git_panel"`.

Also threads the `source` parameter through `activate_draft` and
`ensure_draft` so that drafts created from the sidebar correctly report
`source = "sidebar"` instead of `"agent_panel"`.

Release Notes:

- N/A

Fixes an autoscroll bug that sometimes obscures the current row behind a
sticky header.

The bug was caused by `Editor::autoscroll_vertically` using the
`sticky_header_line_count` value that was cached during the previous
render, which isn't necessarily the number of sticky headers that the
scroll target needs, e.g.
- when jumping to a definition
- when pressing the up arrow key when the selection is in the topmost
visible row with `"vertical_scroll_margin": 0`

This fixes that by not caching `sticky_header_line_count` and instead
querying Tree-sitter on the fly to get the number of sticky headers that
the scroll target needs. Performance-wise this seem okay, I measured it
to take less than 200µs on my machine using a very large Rust file (and
there are still some possible ways to optimize this if necessary). In
particular, the pathological huge single-line file case as discussed in
#48450 shouldn't be an issue here because unlike the main sticky header
Tree-sitter query, here we query the outline items that contain a
particular point rather than those that intersect an entire row.

The `ScrollCursorTop` handler is also simplified to use the same shared
function for counting the number of sticky headers, since that action
doesn't rely on autoscroll.

Before:


https://github.com/user-attachments/assets/efb12776-82d9-4b94-baa5-347ec769fb98

After:


https://github.com/user-attachments/assets/236deb9f-fe06-43bd-b167-7bd3ab719e4c

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 #50803 

Release Notes:

- Fixed sticky headers sometimes obscuring the current row.

---------

Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>

Self-Review Checklist:

- [x] I've reviewed my own diff for quality, security, and reliability
- [ ] Unsafe blocks (if any) have justifying comments
- [x] The content is consistent with the UI/UX checklist
- [x] Tests cover the new/changed behavior
- [x] Performance impact has been considered and is acceptable

Closes #52352

Release Notes:

- Fixed system clipboard being overwritten when pasting over a visual
selection with Cmd-V/Ctrl-V in vim mode