Release Notes:

- Fixed a bug that caused `Reject All` to not always reject _all_ the
hunks.

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

This PR adds a new field in the `MarkdownStyle` struct,
`heading_level_styles`, allowing, via the newly added function
`apply_heading_style` and struct `HeadingLevelStyles` to customize each
individual heading level in Markdown rendering/styling function.

Things like this should now be possible:

```rust
    MarkdownStyle {
        heading_level_styles: Some(HeadingLevelStyles {
            h1: Some(TextStyleRefinement {
                font_size: Some(rems(1.15).into()),
                ..Default::default()
            }),
        }),
        ..Default::default()
    }
```

Release Notes:

- N/A

WIP

<img width="644" alt="image"
src="https://github.com/user-attachments/assets/b24e1a57-f82e-457c-b788-1b314ade7c84"
/>


<img width="644" alt="image"
src="https://github.com/user-attachments/assets/b158953c-2015-4cc8-b8ed-35c6fcbe162d"
/>


Release Notes:

- agent: Improve compatibility with Gemini Tool Calling APIs. When a
tool is incompatible with the Gemini APIs a warning indicator will be
displayed. Incompatible tools will be automatically excluded from the
conversation

---------

Co-authored-by: Danilo Leal <daniloleal09@gmail.com>

If you wanted to interrupt the current LLM response that's generating to
send a follow up message, you'd need to stop it first, type your new
message, and then send it. Now, you can just type your new message while
there's a response generating and send it. This will interrupt the
previous response generation and kick off a new one.

Release Notes:

- agent: Allow to send a new message while a response is generating,
interrupting the LLM to focus instead on the most recent prompt.

This is just a refactor which adds no functionality.
We now return a `ToolResult` from `Tool > run(...)`. For now this just
wraps the output task in a struct. We'll use this to implement custom
rendering of tools, see #28621.

Release Notes:

- N/A

Closes #28638

This PR fixes markdown selection for the info and diagnostic popovers.

In the editor popover, after the changes in
https://github.com/zed-industries/zed/pull/28255, the markdown selection
state updates correctly, but it no longer triggers the editor element to
repaint like it used to. This is fixed by adding a subscription to
listen for markdown entity changes and triggering a repaint for the
editor.

I assume markdown selection works elsewhere because:

1. Either the `Markdown` entity is directly part of a struct that
implements the `Render` trait, causing it to repaint whenever the
markdown state changes. See
[here](https://github.com/zed-industries/zed/blob/d1ffda9bfeccfdf9bea3f76251350bf9cf7f6e1b/crates/ui_prompt/src/ui_prompt.rs#L65).
2. OR it's wrapped around component like Popover which implements
`RenderOnce` trait. See
[here](https://github.com/zed-industries/zed/blob/d1ffda9bfeccfdf9bea3f76251350bf9cf7f6e1b/crates/editor/src/code_context_menus.rs#L645).

Whereas info and diagnostic popovers does not do both. I do think we can
change it to use `Popover` component, but for now this works as quick
fix.

Extras:
- Remove unnecessary struct cloning.
- Refactor rendering logic to use `when_some`.

Release Notes:

- Fixed issue where selection wasn't working for info and diagnostic
popovers.

Adjust the heading sizes for the Agent Panel so they're not aggressively
huge.

Release Notes:

- N/A

Motivation is to emit events when enabled tools change, want to use this
in #28755

Release Notes:

- N/A

WIP

Release Notes:

- agent: Improve compatibility when using MCPs with Gemini models

This is a combination of the "read file" and "list directory contents"
tools as part of a push to reduce our quantity of builtin tools by
combining some of them.

The functionality is all there for this tool, although there's room for
improvement on the visuals side: it currently always shows the same icon
and always says "Read" - so you can't tell at a glance when it's reading
a directory vs an individual file. Changing this will require a change
to the `Tool` trait, which can be in a separate PR. (FYI @danilo-leal!)

<img width="606" alt="Screenshot 2025-04-14 at 11 56 27 PM"
src="https://github.com/user-attachments/assets/bded72af-6476-4469-97c6-2f344629b0e4"
/>

Release Notes:

- Added `contents` tool

Cherry-picked gemini: Fix "invalid argument" error when request contains
no tools (#28747)

When we do not have any tools, we want to set the `tools` field to
`None`

Release Notes:

- Fixed an issue where Gemini requests would sometimes return a Bad
Request ("Invalid argument...")

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

Fine-tuning some areas of the Agent Panel design.

Release Notes:

- N/A

Release Notes:

- N/A

Closes #28475

Updates `rename` and `code_action` `input_schema` methods to use
`json_schema_for<T>()` which transforms standard JSONSchema into the
subset required by Gemini.
Also makes `input_schema` implementations consistent.
Tested tools against Gemini 2.5 Pro Preview, Zed Claude 3.7 Sonnet
Thinking, o3-mini

Release Notes:

- Agent Beta: Fixed error 400 `INVALID_ARGUMENT` when using Gemini with
`code_actions` or `rename` tools enabled.

Release Notes:

- Add support for OpenAI GPT-4.1 via Copilot Chat and OpenAI API

---------

Co-authored-by: Danilo Leal <daniloleal09@gmail.com>
Co-authored-by: Bennet Bo Fenner <bennetbo@gmx.de>

This ensures that we respect the `LanguageModelToolSchemaFormat` value
when we call `tool.input_schema`. This prevents us from breaking Gemini
compatibility when adding/changing built-in tools. See #28634.

The test suite will now fail with an error message like this, when
providing an incompatible input_schema:

```
thread 'tests::test_tool_schema_compatibility' panicked at crates/assistant_tools/src/assistant_tools.rs:108:17:
Tool schema for `code_actions` is not compatible with `language_model::LanguageModelToolSchemaFormat::JsonSchemaSubset` (Gemini Models).
Are you using `schema::json_schema_for<T>(format)` to generate the schema?
```


Release Notes:

- N/A

Ended up not making this one visible only upon hover or something
because the layout alignment would be weird given the list item spans
the full width. So, experimenting with this design here:

<img
src="https://github.com/user-attachments/assets/62bf661e-1aae-4644-8a89-49cefb3e8130"
width="700" />

Release Notes:

- agent: Add scrollbar to the history view.

![CleanShot 2025-04-14 at 11 15
38@2x](https://github.com/user-attachments/assets/9e803ffb-74fd-486b-bebc-2155a407a9fa)

Release Notes:

- agent: Handle context window exceeded errors from Anthropic

This PR cleans up the thread auto-capturing added in #28271.

- Removed usage of `unsafe`
- Fixed feature flag check
- We were incorrectly not respecting the feature flag in release builds
- Made sure the telemetry event was being run on the background executor

Release Notes:

- N/A

Release Notes:

- N/A

When you hit the back button in the agent panel toolbar, we were
returning the focus to the buffer instead to the panel's message editor,
which is likely where you want to be after quickly checking history or
settings.

Release Notes:

- N/A

Allow expanding the message editor while the agent panel is not focused,
right now there is no effect when you use the button from another focus

Release Notes:

- N/A

<img
src="https://github.com/user-attachments/assets/2cdc5121-dd7b-4f46-8d43-88d5152c77ea"
width="550" />


Release Notes:

- N/A

It was too tiny, felt like we could use more breathing room.

Release Notes:

- N/A

Mentioning "MCP" more prominently, adding tool descriptions in the icon
button tooltip, and other UI adjustments.

<img
src="https://github.com/user-attachments/assets/e021b3be-99b8-454c-b5fd-0221a7947a35"
width="600" />

Release Notes:

- N/A

Also took the opportunity to rename the "Continue in New Thread" item to
a potentially clearer name.

<img
src="https://github.com/user-attachments/assets/024de07d-f215-4c41-8fbe-652a216b61d9"
width="300"/>

Release Notes:

- N/A

This PR splits up the rendering of the message editor into multiple
functions. Previously we had a single `h_flex()...` expression which
spanned across 550 lines, `cargo fmt` stopped working.

Release Notes:

- N/A

Release Notes:

- N/A

This code was used before we had a proper completion menu for
at-mentions

Release Notes:

- N/A

Release Notes:

- agent: Show recommended models in the agent model selector and display
the provider in the model selector's trigger.

---------

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

Release Notes:

- agent: Start language servers when accessing files via tools

Co-authored-by: Michael <michael@zed.dev>

Closes #ISSUE

Release Notes:

- Fixed a bug that could cause the staged status of entries in the git
panel to be stale.

Cherry-picked Try to weak-link ScreenCaptureKit always (#28585)

Release Notes:

- Fixed a crash on macOS (Catalina, Big Sur). Users may need to manually
redownload Zed if they updated to a broken release.
([#28585](https://github.com/zed-industries/zed/pull/28585))

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

Cherry-picked Fix a panic in the git store (#28590)

Closes #ISSUE

Release Notes:

- Fixed a panic that could occur when git statuses were updated.

Co-authored-by: Cole Miller <cole@zed.dev>

Release Notes:

- agent: Fix a bug where an inserted @mention did not show up as the one
that was selected

Moving from "NewPromptEditor" to "NewTextThread". We recently re-named
that and this was missing.

Release Notes:

- N/A

This PR allows expanding the message editor textarea to fit almost the
total height of the Agent Panel. Stylistically, I'm also changing the
font family we use in the textarea to use the buffer font; want to
experiment with this for a bit.

Release Notes:

- agent: The Agent Panel textarea can now be expanded to fill almost the
total height of the panel.

---------

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

This PR lightly refactors the `EditorMode::Full` exposing two new
methods: `is_full` and `set_mode`.

Motivation is to expose fields that modify the behavior when the editor
is in `Full` mode. By using is `mode.is_full()` instead of
`EditorMode::Full` we can introduce new fields without breaking other
places in the code.

Release Notes:

- N/A

Co-authored-by: Bennet Bo Fenner <bennetbo@gmx.de>

Release Notes:

- Fixed a regression that caused the agent to hang sometimes.

---------

Co-authored-by: Thomas Mickley-Doyle <tmickleydoyle@gmail.com>
Co-authored-by: Nathan Sobo <nathan@zed.dev>
Co-authored-by: Michael Sloan <mgsloan@gmail.com>

This allows us to not scan the codeblock content for newlines on every
frame in `active_thread`

Release Notes:

- N/A

From Claude 3.5 Sonnet to **Claude 3.7 Sonnet**.

Release Notes:

- Change the default model of Zed's hosted LLM service to Claude 3.7
Sonnet.

This PR fixes usage of non-streaming Copilot Chat models.

Closes https://github.com/zed-industries/zed/issues/28528.

Release Notes:

- Fixed an issue with using non-streaming Copilot Chat models (e.g., o1,
o3-mini).