This PR adds a new WebSocket connection to Cloud.
This connection will be used to push down notifications from the server
to the client.
Release Notes:
- N/A
---------
Co-authored-by: Marshall Bowers <git@maxdeviant.com>
Now the edit tool can access files outside the current project (just
like the terminal tool can), but it's behind a prompt (unlike other edit
tool actions).
Release Notes:
- The edit tool can now access files outside the current project, but
only if the user grants it permission to.
Someone encountered this in production, which should not happen:
<img width="1266" height="623" alt="Screenshot 2025-07-24 at 10 38
40 AM"
src="https://github.com/user-attachments/assets/40f3f977-5110-4808-a456-7e708d953b3b"
/>
This moves certain errors into the category of "never retry" and reduces
the number of retries for some others. Also it adds some diagnostic
logging for retry policy.
It's not a complete fix for the above, because the underlying issues is
that the server is sending a HTTP 403 response and although we were
already treating 403s as "do not retry" it was deciding to retry with 2
attempts anyway. So further debugging is needed to figure out why it
wasn't going down the 403 branch by the time the request got here.
Release Notes:
- N/A
Now we explicitly carve out exceptions for which HTTP responses we do
*not* retry for, and retry at least once on all others.
Release Notes:
- The Agent panel now automatically retries failed requests under more
circumstances.
Now we only auto-retry if burn mode is enabled. We also show a "Retry"
button (so you don't have to type "continue") if you think that's the
right remedy, and additionally we show a "Retry and Enable Burn Mode"
button if you don't have it enabled.
<img width="484" height="260" alt="Screenshot 2025-07-17 at 6 25 27 PM"
src="https://github.com/user-attachments/assets/dc5bf1f6-8b11-4041-87aa-4f37c95ea9f0"
/>
<img width="478" height="307" alt="Screenshot 2025-07-17 at 6 22 36 PM"
src="https://github.com/user-attachments/assets/1ed6578a-1696-449d-96d1-e447d11959fa"
/>
Release Notes:
- Only auto-retry Agent requests when Burn Mode is enabled
In #33275 I was very conservative about when to retry when there are
errors in language completions in the Agent panel.
Now we retry in more scenarios (e.g. HTTP 5xx and 4xx errors that aren't
in the specific list of ones that we handle differently, such as 429s),
and also we show a notification if the thread halts for any reason.
<img width="441" height="68" alt="Screenshot 2025-07-15 at 12 51 30 PM"
src="https://github.com/user-attachments/assets/433775d0-a8b2-403d-9427-1e296d164980"
/>
<img width="482" height="322" alt="Screenshot 2025-07-15 at 12 44 15 PM"
src="https://github.com/user-attachments/assets/5a508224-0fe0-4d34-9768-25d95873eab8"
/>
Release Notes:
- Automatic retry for more Agent errors
- Whenever the Agent stops, play a sound (if configured) and show a
notification (if configured) if the Zed window was in the background.
It turns out Starship is using custom Powerline separators in the
Unicode private reserved character range. This addresses some issues
seen in the comments of #34234
Release Notes:
- Fix automatic contrast adjustment for Powerline separators
Closes#33253 in a way that doesn't regress #32175 - namely,
automatically adjusts the contrast between the foreground and background
text in the terminal such that it's above a certain threshold. The
threshold is configurable in settings, and can be set to 0 to turn off
this feature and use exactly the colors the theme specifies even if they
are illegible.
## One Light Theme Before
<img width="220" alt="Screenshot 2025-07-07 at 6 00 47 PM"
src="https://github.com/user-attachments/assets/096754a6-f79f-4fea-a86e-cb7b8ff45d60"
/>
(Last row is highlighted because otherwise the text is unreadable; the
foreground and background are the same color.)
## One Light Theme After
(This is with the new default contrast adjustment setting.)
<img width="215" alt="Screenshot 2025-07-07 at 6 22 02 PM"
src="https://github.com/user-attachments/assets/b082fefe-76f5-4231-b704-ff387983a3cb"
/>
This approach was inspired by @mitchellh's use of automatic contrast
adjustment in [Ghostty](https://ghostty.org/) - thanks, Mitchell! The
main difference is that we're using APCA's formula instead of WCAG for
[these
reasons](https://khan-tw.medium.com/wcag2-are-you-still-using-it-ui-contrast-visibility-standard-readability-contrast-f34eb73e89ee).
Release Notes:
- Added automatic dynamic contrast adjustment for terminal foreground
and background colors
Closes#33761
The problem was that in the indentation regex we were treating lines
that had `:` in them as requiring an indent on the next line, even if
that `:` was inside a comment.
Release Notes:
- Fixed YAML indentation for lines containing comments with `:` in them
<img width="484" alt="Screenshot 2025-06-25 at 2 26 16 PM"
src="https://github.com/user-attachments/assets/340f15d7-b115-4895-bae8-b12a915bfda1"
/>
<img width="460" alt="Screenshot 2025-06-25 at 2 26 08 PM"
src="https://github.com/user-attachments/assets/6e587a38-d542-405f-809f-402e87520538"
/>
Now we:
* Automatically retry up to 3 times on upstream Overloaded or 500 errors
(currently for Anthropic only; will add others in future PRs)
* Also automatically retry on rate limit errors (using the provided
duration to wait, if we were given one)
* Give you a notification if you don't have Zed open and we stopped the
thread because of an error
Still todo in future PRs:
* Update collab to report Overloaded and 500 errors differently if
collab itself is passing through an upstream error vs not (currently we
report these as "Zed's API is overloaded" when actually it's the
upstream one!)
* Updating providers other than Anthropic to categorize their errors so
that they benefit from this
* Expanding graceful error handling/retry to other things besides
Overloaded and 500 errors (e.g. connection reset)
Release Notes:
- Automatically retry in Agent Panel instead of erroring out when an
upstream AI API is overloaded or 500s
- Show a notification when an Agent thread errors out and Zed is not the
active window
While working on retries, I discovered some opportunities to reduce
cloning of message segments. These segments have full `String`s (not
`SharedString`s), so cloning them means copying cloning all the bytes of
all the strings in the message, which would be nice to avoid!
Release Notes:
- N/A
This PR is in preparation for doing automatic retries for certain
errors, e.g. Overloaded. It doesn't change behavior yet (aside from some
granularity of error messages shown to the user), but rather mostly
changes some error handling to be exhaustive enum matches instead of
`anyhow` downcasts, and leaves some comments for where the behavior
change will be in a future PR.
Release Notes:
- N/A
Previously we were using a mix of `u32` and `usize`, e.g. `max_tokens:
usize, max_output_tokens: Option<u32>` in the same `struct`.
Although [tiktoken](https://github.com/openai/tiktoken) uses `usize`,
token counts should be consistent across targets (e.g. the same model
doesn't suddenly get a smaller context window if you're compiling for
wasm32), and these token counts could end up getting serialized using a
binary protocol, so `usize` is not the right choice for token counts.
I chose to standardize on `u64` over `u32` because we don't store many
of them (so the extra size should be insignificant) and future models
may exceed `u32::MAX` tokens.
Release Notes:
- N/A
Minor refactor that I'm extracting from a branch because it can stand
alone.
- Now we no longer spawn an executor for `report_anthropic_event` if
it's just going to immediately fail due to API key being missing
- `report_anthropic_event` now takes a `String` API key instead of
`Option<String>` and the error reporting if the key is missing has been
moved to the caller.
- `report_anthropic_event` is longer coupled to `AnthropicError`,
because all it ever did was generate an `AnthropicEvent::Other`, which
in turn was then only used for `log_err` - so, can just be an
`anyhow::Result`.
Release Notes:
- N/A
Based on a Slack conversation with @notpeter - this prevents secrets in
private/excluded files from being sent by the agent to third parties for
tools that don't require confirmation.
Of course, the agent can still use the terminal tool or MCP to access
these, but those require confirmation before they run (unlike these
tools).
This change doesn't seem to cause any trouble for evals:
<img width="730" alt="Screenshot 2025-06-03 at 8 48 33 PM"
src="https://github.com/user-attachments/assets/d90221be-f946-4af2-b57b-4aa047e86853"
/>
Release Notes:
- N/A
Re-enables format on save for agent changes (when the user has that
enabled in settings), except differently from before:
- Now we do the format-on-save in the separate buffer the edit tool
uses, *before* the diff
- This means it never triggers separate staleness
- It has the downside that edits are now blocked on the formatter
completing, but that's true of saving in general.
Release Notes:
- N/A
---------
Co-authored-by: Agus Zubiaga <hi@aguz.me>
This expands our deserialization of JSON from models to be more tolerant
of different variations that the model may send, including
capitalization, wrapping things in objects vs. being plain strings, etc.
Also when deserialization fails, it reports the entire error in the JSON
so we can see what failed to deserialize. (Previously these errors were
very unhelpful at diagnosing the problem.)
Finally, also removes the `WrappedText` variant since the custom
deserializer just turns that style of JSON into a normal `Text` variant.
Release Notes:
- N/A
This is a partial reversion of
https://github.com/zed-industries/zed/pull/30878 - having it always
scroll to bottom whenever a new message is added makes it so that when
you're scrolled up, you don't have time to read what you're trying to
read before it autoscrolls to the end.
@danilo-leal when you're back, we can pair on addressing that in a
different way!
Release Notes:
- Fixed bug where scrolling up in the agent panel didn't prevent
automatic scroll-to-end whenever a new message arrived.
Some providers sometimes send `{ "type": "text", "text": ... }` instead
of just the text as a string. Now we accept those instead of erroring.
Release Notes:
- N/A
This is very basic support for them. There are a number of other TODOs
before this is really a first-class supported feature, so not adding any
release notes for it; for now, this PR just makes it so that if
read_file tries to read a PNG (which has come up in practice), it at
least correctly sends it to Anthropic instead of messing up.
This also lays the groundwork for future PRs for more first-class
support for images in tool calls across more image file formats and LLM
providers.
Release Notes:
- N/A
---------
Co-authored-by: Agus Zubiaga <hi@aguz.me>
Co-authored-by: Agus Zubiaga <agus@zed.dev>
This is our first eval of the Minimal tool profile. Right now they're
all passing; the value of having it is to catch regressions in the
system prompt (which has special logic in it for the case where no tools
are enabled).
Release Notes:
- N/A
Sometimes models return absolute paths even though we ask them not to
(including sometimes returning `/dev/null`). Currently we assume we're
always given a relative path, which leads to a panic in debug builds.
Now we just support being given absolute paths.
Release Notes:
- N/A
---------
Co-authored-by: Agus Zubiaga <hi@aguz.me>
Also `now` in `write` profile
Release Notes:
- Tools for manipulating directories no longer require confirmation, and
are enabled in the Write profile
- Enabled `now` and `list_directory` tools by default in Write profile
---------
Co-authored-by: Cole Miller <m@cole-miller.net>
Co-authored-by: Cole Miller <cole@zed.dev>
Release Notes:
- Fixed a bug where clicking context pills switched into the "editing
message" state instead of clicking the pill.
Co-authored-by: Michael <michael@zed.dev>
Co-authored-by: Ben <ben@zed.dev>
This is desirable for when we want to use a `ButtonLike` to show a
tooltip over an icon, and we don't want it to show the "not allowed"
cursor on hover.
Release Notes:
- N/A
I saw a slice panic (for begin > end) in a debug build of the eval. This
should just be a failed assertion, not a panic that takes out the whole
eval run!
Release Notes:
- N/A
This is based on having observed that there is a lot of variation
between runs on `n=1` and `n=3`.
* With `n=8` two runs on the same branch give answers that seem close
enough to be reasonably consistent.
* With higher concurrency, trying to run this many repetitions seems to
lead language servers to time out a lot, causing evals to fail.
Release Notes:
- N/A
Fixed issue where eval thread judges were not considering the last
response in the thread.
The problem was that they were getting the full list of messages from
`last_request`, which (being a request!) did not have the response yet.
Release Notes:
- N/A
Now we're more tolerant of invalid JSON coming back from the model
(possibly because it was incomplete and we're streaming), plus if we do
end up with invalid JSON once it has all streamed back, we report what
the malformed JSON actually was:
<img width="444" alt="Screenshot 2025-04-23 at 1 49 14 PM"
src="https://github.com/user-attachments/assets/480f5da7-869b-49f3-9ffd-8f08ccddb33d"
/>
Release Notes:
- N/A
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
Having a separate rename tool seems to make the agent more likely to use
it compared to having it be part of the code actions tool.
Release Notes:
- Added code action tool and rename tool.
This reverts commit f12a554f86, which
introduced a panic in inline assistant (cc @mgsloan) - I'm not sure what
the motivation was for that change, but I figure we can revert to fix
the inline assistant now and deal with that later. 😄
Panic was:
> Thread "main" panicked with "cannot read workspace::Workspace while it
is already being updated" at
/Users/rtfeldman/code/zed/crates/gpui/src/app/entity_map.rs:139:32
Release Notes:
- N/A
Trying out sending the model a reminder message about code blocks in the
system prompt. If this seems to work well, we can include more specific
reminder messages, e.g. tool-specific ones.
Release Notes:
- N/A
Sometimes agents do this. I've had some success responding by telling it
not to do this, so trying out having it in the system prompt.
Release Notes:
- Adjusted the system prompt to avoid incomplete code generation.
Currently, it's pretty common that when the agent gets stuck, it deletes
whatever it's stuck on and replaces it with a TODO comment, then
cheerfully reports that it has "simpified" the implementation. This is
worse than leaving the broken code, because at least a human could take
over and try to get it across the finish line.
This system prompt adjustment attempts to make the agent do something
more useful when in this situation: report that it's stuck, explain why
it's stuck, and ask the user what to do.
Release Notes:
- N/A
<img width="622" alt="Screenshot 2025-04-05 at 5 48 14 PM"
src="https://github.com/user-attachments/assets/24b9c7d4-d3e2-4929-bca8-79db5b4e5748"
/>
Release Notes:
- The `read_files` tool now reads only the symbol outline files above a
certain size, to conserve context window space. Then it suggests that
the agent call `read_files` again with the relevant line ranges it saw
in the outline.
Lets you get all the code symbols in the project (like the Code Symbols
panel) or in a particular file (like the Outline panel), optionally
paginated and filtering results by regex. The tool gives the files,
lines, and numbers of all of these, which means they can be used in
conjunction with the read file tool to read subsets of large files
without having to open the entire large file and poke around in it.
<img width="621" alt="Screenshot 2025-03-29 at 12 00 21 PM"
src="https://github.com/user-attachments/assets/d78259d7-2746-44c0-ac18-2e21f2505c0a"
/>
Release Notes:
- N/A
By default, agent notifications now display only on your primary screen.
You can optionally configure them to display on all screens (or not to
display at all).
Release Notes:
- N/A
This caused undesirable and unnecessary scrolling, in that if you scroll
up, then as new messages stream into the panel, they jump the scroll bar
back down.
@agu-z and I paired on this and we think this is unnecessary now that we
don't see the Edit button in the UI, but @bennetbo we still see code for
the button in there, so...maybe we do still want this? (If so, we can
revert the second commit and go back to a more conditional way of
scrolling. 😄)
Release Notes:
- N/A
---------
Co-authored-by: Agus <agus@zed.dev>
<img width="620" alt="Screenshot 2025-03-27 at 2 29 13 PM"
src="https://github.com/user-attachments/assets/dd023507-61bc-4722-a095-f65f4b6c746a"
/>
We'll iterate on the UI, but first the goal is to just get it to work at
all so we can see if it's useful in terms of getting correct output
faster.
Release Notes:
- N/A
---------
Co-authored-by: Agus Zubiaga <hi@aguz.me>
Now if you refocus Zed manually (e.g. cmd-tab), we hide the "View Panel"
notification automatically.
This also fixes a related subscription leak.
Release Notes:
- N/A
@agu-z and paired on trying out a "one tool call per edit" approach for
editing files. (The previous approach is still available, it's just
unchecked by default for now.)
Release Notes:
- N/A
---------
Co-authored-by: Agus <agus@zed.dev>
Sometimes we've seen models provide an empty string for the path search
glob. This assumes they meant "*" when that happens.
Separately, this also removes an unnecessary `clone` of a `String`.
Release Notes:
- N/A
We decided to take this out for now. It doesn't seem necessary, and it
complicates the code a lot. We can always put it back later if desired.
Release Notes:
- N/A
If the tool asks to read a path, we don't need to verify whether that
path exists on disk; an unsaved buffer with that path is fine.
Release Notes:
- N/A
Richard Feldman
