Nathan here: I also tacked on a bunch of UI refinement.
Release Notes:
- Introduced the ability to follow the agent around as it reads and
edits files.
---------
Co-authored-by: Nathan Sobo <nathan@zed.dev>
Co-authored-by: Max Brunsfeld <maxbrunsfeld@gmail.com>
This PR adds a notice when reaching consecutive tool use limits when
using normal mode.
Here's an example with the limit artificially lowered to 2 consecutive
tool uses:
https://github.com/user-attachments/assets/32da8d38-67de-4d6b-8f24-754d2518e5d4
Release Notes:
- agent: Added a notice when reaching consecutive tool use limits when
using a model in normal mode.
- [x] Separate MCP servers from tools in the profile customization modal
view
- [x] Group MCP tools in the MCP picker and add a heading
- [x] Separate bult-in profiles from custom ones in the dropdown
selector
- [x] Separate bult-in profiles from custom ones in the modal
- [ ] Enable looping through items via keybinding without opening the
dropdown (will be done on a follow-up PR)
- [ ] Stretch: Focus on the currently active item upon opening the
dropdown (will be done on a follow-up PR)
Release Notes:
- N/A
---------
Co-authored-by: Bennet Bo Fenner <53836821+bennetbo@users.noreply.github.com>
Co-authored-by: Bennet Bo Fenner <bennetbo@gmx.de>
Co-authored-by: Bennet Bo Fenner <bennet@zed.dev>
Follow-up of https://github.com/zed-industries/zed/pull/29706
Instead of doing `cargo check` manually, use rust-analyzer's flycheck:
at the cost of more sophisticated check command configuration, we keep
much less code in Zed, and get a proper progress report.
User-facing UI does not change except `diagnostics_fetch_command` and
`env` settings removed from the diagnostics settings.
Release Notes:
- N/A
This pull request introduces a new tool for streaming edits. The
short-term goal is for this tool to replace the existing `EditFileTool`,
but we want to get this out the door as soon as possible so that we can
start testing it.
`StreamingEditFileTool` is mutually exclusive with `EditFileTool`. It
will be enabled by default for anyone who has the `agent-stream-edits`
feature flag, as well as people that set `assistant.stream_edits` to
`true` in their settings.
### Implementation
Streaming is achieved by requesting a completion while the `edit_file`
tool gets called. We invoke the model by taking the existing
conversation with the agent and appending a prompt specifically tailored
for editing. In that prompt, we ask the model to produce a stream of
`<old_text>`/`<new_text>` tags. As the model streams text in, we
incrementally parse it and start editing as soon as we can.
### Evals
Note that, as part of this pull request, I also defined some new evals
that I used to drive the behavior of the recursive LLM call. To run
them, use this command:
```bash
cargo test --package=assistant_tools --features eval -- eval_extract_handle_command_output
```
Or comment out the `#[cfg_attr(not(feature = "eval"), ignore)]` macro.
I recommend running them one at a time, because right now we don't
really have a way of orchestrating of all these evals. I think we should
invest into that effort once the new agent panel goes live.
Release Notes:
- N/A
---------
Co-authored-by: Nathan Sobo <nathan@zed.dev>
Co-authored-by: Bennet Bo Fenner <bennetbo@gmx.de>
Co-authored-by: Oleksiy Syvokon <oleksiy.syvokon@gmail.com>
Adjusts the way `cargo` and `rust-analyzer` diagnostics are fetched into
Zed.
Nothing is changed for defaults: in this mode, Zed does nothing but
reports file updates, which trigger rust-analyzers'
mechanisms:
* generating internal diagnostics, which it is able to produce on the
fly, without blocking cargo lock.
Unfortunately, there are not that many diagnostics in r-a, and some of
them have false-positives compared to rustc ones
* running `cargo check --workspace --all-targets` on each file save,
taking the cargo lock
For large projects like Zed, this might take a while, reducing the
ability to choose how to work with the project: e.g. it's impossible to
save multiple times without long diagnostics refreshes (may happen
automatically on e.g. focus loss), save the project and run it instantly
without waiting for cargo check to finish, etc.
In addition, it's relatively tricky to reconfigure r-a to run a
different command, with different arguments and maybe different env
vars: that would require a language server restart (and a large project
reindex) and fiddling with multiple JSON fields.
The new mode aims to separate out cargo diagnostics into its own loop so
that all Zed diagnostics features are supported still.
For that, an extra mode was introduced:
```jsonc
"rust": {
// When enabled, Zed runs `cargo check --message-format=json`-based commands and
// collect cargo diagnostics instead of rust-analyzer.
"fetch_cargo_diagnostics": false,
// A command override for fetching the cargo diagnostics.
// First argument is the command, followed by the arguments.
"diagnostics_fetch_command": [
"cargo",
"check",
"--quiet",
"--workspace",
"--message-format=json",
"--all-targets",
"--keep-going"
],
// Extra environment variables to pass to the diagnostics fetch command.
"env": {}
}
```
which calls to cargo, parses its output and mixes in with the existing
diagnostics:
https://github.com/user-attachments/assets/e986f955-b452-4995-8aac-3049683dd22c
Release Notes:
- Added a way to get diagnostics from cargo and rust-analyzer without
mutually locking each other
- Added `ctrl-r` binding to refresh diagnostics in the project
diagnostics editor context
This moves summarization task management out of `context_store`. The
code there was draining a Vec of tasks to block on, but this is no
longer a good fit for message_editor's context loading. It needs to be
able to repeatedly await on the thread summarization tasks involved in
the context.
Discussed with Danilo, and he thinks it'd be good to remove the current
"Summarizing context" anyway since it causes layout shift. If message
send is blocked on summarizing, the pulsing context pill is sufficient
for now. This UI change made this overall change more straightforward.
Release Notes:
- N/A
This change updates the system prompt to conditionally include
`grep`-related instructions based on whether the `grep` tool is enabled.
Implementation details:
1. Add a `has_tool` handlebars helper.
2. Pass the `model` to all locations where the prompt is built.
3. Use `{{#if has_tool "grep"}}` in the system prompt to gate
`grep`-specific instructions.
Testing:
- Unit tests for the `hasTool` helper.
- Unit tests to verify that `grep`-related instructions are included /
omitted from the prompt as appropriate.
- Manual agent evaluation:
- Setup: Asked the Agent "List all impls of MyTrait in the project"
using a custom "No tools" profile (all tools disabled).
- Before the change: The Agent attempted to call `grep`, encountered an
error, then realized the tool was unavailable.
- After the change: The Agent immediately asked to enable a search tool.
Note: in principle, `grep`/`read_file` tool descriptions alone might be
enough, but to confirm this we need more evaluation. If it turns out to
be true, we'll be able to remove grep-specific instructions from the
system prompt and undo this change.
Release Notes:
- N/A
Closes#27641
This PR fixes invalid proxy URIs being registered despite the URI not
being a valid proxy URI.
Whilst investigating #27641 , I noticed that currently any proxy URI
passed to `RequestClient::proxy_and_user_agent` will be assigned to the
created client, even if the URI is not a valid proxy URI. Given a test
as an example:
We create an URI here and pass it as a proxy to
`ReqwestClient::proxy_and_user_agent`:
https://github.com/zed-industries/zed/blob/main/crates/reqwest_client/src/reqwest_client.rs#L272-L273
In `ReqwestClient::proxy_and_user_agent`we take the proxy parameter here
9b40770e9f/crates/reqwest_client/src/reqwest_client.rs (L46)
and set it unconditionally here:
9b40770e9f/crates/reqwest_client/src/reqwest_client.rs (L62)
, not considering at all whether the proxy was successfully created
above. Concluding, we currently do not actually check whether a proxy
was successfully created, but rather whether an URI is equal to itself,
which trivially holds. The existing test for a malformed proxy URI
9b40770e9f/crates/reqwest_client/src/reqwest_client.rs (L293-L297)
does not check whether invalid proxies cause an error, but rather checks
whether `http::Uri::from_static` panics on an invalid URI, [which it
does as
documented](https://docs.rs/http/latest/http/uri/struct.Uri.html#panics).
Thus, the tests currently do not really check anything proxy-related and
invalid proxies are assigned as valid proxies.
---
This PR fixes the behaviour by considering whether the proxy was
actually properly parsed and only assigning it if that is the case.
Furthermore, it improves logging in case of errors so issues like the
linked one are easier to debug (for the linked issue, the log will now
include that the proxy schema is not supported in the logs).
Lastly, it also updates the test for a malformed proxy URI. The test now
actually checks that malformed proxy URIs are not registered for the
client rather than testing the `http` crate.
The update also initially caused the [test for a `socks4a`
proxy](9b40770e9f/crates/reqwest_client/src/reqwest_client.rs (L280C1-L282C50))
to fail. This happened because the reqwest-library introduced supports
for `socks4a` proxies in [version
0.12.13](https://github.com/seanmonstar/reqwest/blob/master/CHANGELOG.md#v01213).
Thus, this PR includes a bump of the reqwest library to add proper
support for socks4a proxies.
Release Notes:
- Added support for socks4a proxies.
---------
Co-authored-by: Peter Tripp <peter@zed.dev>
resolves#24655resolves#23945
I haven't yet added a default binding for the new command. #27797 added `:ls` and
`:buffers` which in my opinion should use the global searchable version
given that that matches the vim semantics of those commands better than
just showing the tabs in the local pane.
There's also a question of what to do when you select a tab from another
pane, should the focus jump to that pane or should that tab move to the
currently focused pane? For now I've implemented the former.
Release Notes:
- Added `tab_switcher::ToggleAll` to search open tabs from all panes and focus the selected one.
---------
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
Closes#29439
Add shell escaping as well as additional sanity check for Zed path when
used in askpass. This caused issues on preview and nightly as the
standard paths for those releases contain spaces which were not escaped
appropriately leading to erroneous "Permission denied" errors from SSH
when the askpass script failed
Release Notes:
- Fixed a missing shell-escape in askpass resulting in erroneous
"Permission denied" errors when trying to connect to a remote server
over ssh (effecting preview release v0.184.1 and nightly only)
Implementing the `ToolCard` for the path_search tool. It also adds the
"jump to file" functionality if you expand the results.
Release Notes:
- N/A
---------
Co-authored-by: Richard Feldman <oss@rtfeldman.com>
Co-authored-by: Agus Zubiaga <hi@aguz.me>
This PR adds support for the eval to read environment variables from a
`.env` file located in the `crates/eval` directory.
For instance, you can use it to set your Anthropic API key:
```
ANTHROPIC_API_KEY=<secret>
```
Release Notes:
- N/A
This update generates a single self-contained .html file that shows an
overview of evaluation threads in the browser. It's useful for:
- Quickly reviewing results
- Sharing evaluation runs
- Debugging
- Comparing models (TBD)
Features:
- Export thread JSON from the UI
- Keyboard navigation (j/k or Ctrl + ←/→)
- Toggle between compact and full views
Generating the overview:
- `cargo run -p eval` will write this file in the run dir's root.
- Or you can call `cargo run -p eval --bin explorer` to generate it
without running evals.
Screenshot:
Release Notes:
- N/A
Part of https://github.com/zed-industries/zed/issues/16472
* Adds debug logging to everywhere near INSERT/UPDATEs in the DB
So something like
`env RUST_LOG=debug,wasmtime_cranelift=off,cranelift_codegen=off,vte=off
cargo run` could be used to view these (current zlog seems to process
the exclusions odd, so not sure this is the optimal RUST_LOG line) can
be used to debug any further writes.
* Removes excessive window stack serialization
Previously, it serialized unconditionally every 100ms.
Now, only if the stack had changed, which is now check every 500ms.
* Removes excessive terminal serialization
Previously, it serialized its `cwd` on every `ItemEvent::UpdateTab`
which was caused by e.g. any character output.
Now, only if the `cwd` has changed at the next event processing time.
Release Notes:
- Fixed more excessive DB writes
Simplifies the data structures involved in agent context by removing
caching and limiting the use of ContextId:
* `AssistantContext` enum is now like an ID / handle to context that
does not need to be updated. `ContextId` still exists but is only used
for generating unique `ElementId`.
* `ContextStore` has a `IndexMap<ContextSetEntry>`. Only need to keep a
`HashSet<ThreadId>` consistent with it. `ContextSetEntry` is a newtype
wrapper around `AssistantContext` which implements eq / hash on a subset
of fields.
* Thread `Message` directly stores its context.
Fixes the following bugs:
* If a context entry is removed from the strip and added again, it was
reincluded in the next message.
* Clicking file context in the thread that has been removed from the
context strip didn't jump to the file.
* Refresh of directory context didn't reflect added / removed files.
* Deleted directories would remain in the message editor context strip.
* Token counting requests didn't include image context.
* File, directory, and symbol context deduplication relied on
`ProjectPath` for identity, and so didn't handle renames.
* Symbol context line numbers didn't update when shifted
Known bugs (not fixed):
* Deleting a directory causes it to disappear from messages in threads.
Fixing this in a nice way is tricky. One easy fix is to store the
original path and show that on deletion. It's weird that deletion would
cause the name to "revert", though. Another possibility would be to
snapshot context metadata on add (ala `AddedContext`), and keep that
around despite deletion.
Release Notes:
- N/A
There's probably more to do to fully make the transition, and we'll
still debate a bit internally whether this is the name, but just opening
this PR up now for visibility.
Release Notes:
- N/A
This PR wires the counting of Google AI tokens back up.
It now goes through the LLM service instead of collab's RPC.
Still only available for Zed staff.
Release Notes:
- N/A
This PR removes the `CountLanguageModelTokens` RPC message from collab.
We were only using this for Google AI models through the Zed provider
(which is only available to Zed staff).
For now we're returning `0`, but will bring back soon.
Release Notes:
- N/A
This PR uses Tree Sitter to show inline values while a user is in a
debug session.
We went with Tree Sitter over the LSP Inline Values request because the
LSP request isn't widely supported. Tree Sitter is easy for
languages/extensions to add support to. Tree Sitter can compute the
inline values locally, so there's no need to add extra RPC messages for
Collab. Tree Sitter also gives Zed more control over how we want to show
variables.
There's still more work to be done after this PR, namely differentiating
between global/local scoped variables, but it's a great starting point
to start iteratively improving it.
Release Notes:
- N/A
---------
Co-authored-by: Piotr Osiewicz <peterosiewicz@gmail.com>
Co-authored-by: Anthony Eid <hello@anthonyeid.me>
Co-authored-by: Cole Miller <m@cole-miller.net>
Co-authored-by: Anthony <anthony@zed.dev>
Co-authored-by: Kirill <kirill@zed.dev>
Things this doesn't currently handle:
- [x] ~testing~
- ~we really need an snapshot test that takes a vscode settings file
with all options that we support, and verifies the zed settings file you
get from importing it, both from an empty starting file or one with lots
of conflicts. that way we can open said vscode settings file in vscode
to ensure that those options all still exist in the future.~
- Discussed this, we don't think this will meaningfully protect us from
future failures, and we will just do this as a manual validation step
before merging this PR. Any imports that have meaningfully complex
translation steps should still be tested.
- [x] confirmation (right now it just clobbers your settings file
silently)
- it'd be really cool if we could show a diff multibuffer of your
current settings with the result of the vscode import and let you pick
"hunks" to keep, but that's probably too much effort for this feature,
especially given that we expect most of the people using it to have an
empty/barebones zed config when they run the import.
- [x] ~UI in the "welcome" page~
- we're planning on redoing our welcome/walkthrough experience anyways,
but in the meantime it'd be nice to conditionally show a button there if
we see a user level vscode config
- we'll add it to the UI when we land the new walkthrough experience,
for now it'll be accessible through the action
- [ ] project-specific settings
- handling translation of `.vscode/settings.json` or `.code-workspace`
settings to `.zed/settings.json` will come in a future PR, along with UI
to prompt the user for those actions when opening a project with local
vscode settings for the first time
- [ ] extension settings
- we probably want to do a best-effort pass of popular extensions like
vim and git lens
- it's also possible to look for installed/enabled extensions with `code
--list-extensions`, but we'd have to maintain some sort of mapping of
those to our settings and/or extensions
- [ ] LSP settings
- these are tricky without access to the json schemas for various
language server extensions. we could probably manage to do translations
for a couple popular languages and avoid solving it in the general case.
- [ ] platform specific settings (`[macos].blah`)
- this is blocked on #16392 which I'm hoping to address soon
- [ ] language specific settings (`[rust].foo`)
- totally doable, just haven't gotten to it yet
~We may want to put this behind some kind of flag and/or not land it
until some of the above issues are addressed, given that we expect
people to only run this importer once there's an incentive to get it
right the first time. Maybe we land it alongside a keymap importer so
you don't have to go through separate imports for those?~
We are gonna land this as-is, all these unchecked items at the bottom
will be addressed in followup PRs, so maybe don't run the importer for
now if you have a large and complex VsCode settings file you'd like to
import.
Release Notes:
- Added a VSCode settings importer, available via a
`zed::ImportVsCodeSettings` action
---------
Co-authored-by: Mikayla Maki <mikayla@zed.dev>
Co-authored-by: Kirill Bulatov <kirill@zed.dev>
Co-authored-by: Mikayla Maki <mikayla.c.maki@gmail.com>
Co-authored-by: Marshall Bowers <git@maxdeviant.com>
This PR implements the `ToolCard` for the edit file tool, which allow us
to display an editor with a diff in the thread view with the changes
performed by the model.
- [x] Fix buffer sometimes displaying empty
- [x] Stop buffer from scrolling together with the thread
- [x] Fix multibuffer header sometimes appearing
- [x] Fix buffer height issue
- [x] Implement "full height" expand button
- [x] Add "Jump To File" functionality
- [x] Polish and refine styles
Release Notes:
- agent: Added diff preview cards in the thread view for edits performed
by the agent.
---------
Co-authored-by: João Marcos <marcospb19@hotmail.com>
Co-authored-by: Richard Feldman <oss@rtfeldman.com>
Co-authored-by: Agus Zubiaga <hi@aguz.me>
Co-authored-by: Conrad Irwin <conrad.irwin@gmail.com>
Now all debug sessions are routed through the debug panel and are
started synchronously instead of by a task that returns a session once
the initialization process is finished. A session is `Mode::Booting`
while it's starting the debug adapter process and then transitions to
`Mode::Running` once this is completed.
This PR also added new tests for the dap logger, reverse start debugging
request, and debugging over SSH.
Release Notes:
- N/A
---------
Co-authored-by: Anthony Eid <hello@anthonyeid.me>
Co-authored-by: Anthony <anthony@zed.dev>
Co-authored-by: Cole Miller <m@cole-miller.net>
Co-authored-by: Cole Miller <cole@zed.dev>
Co-authored-by: Zed AI <ai@zed.dev>
Co-authored-by: Remco Smits <djsmits12@gmail.com>
### Problem
We want to start continuously tracking our progress on agent evals over
time. As part of this, we'd like the *score* to have a clear,
interpretable meaning. Right now, it's a number from 0 to 5, but it's
not clear what any particular number works. In addition, scores vary
widely from run to run, because the agent's output is deterministic. We
try to stabilize the score using a panel of judges, but the behavior of
the agent itself varies much more widely than the judges' scores for a
given run.
### Solution
* **explicit meanings of scores** - In this PR, we're prescribing the
diff and thread criteria files so that they *must* be unordered lists of
assertions. For both the thread and the diff, rather than providing an
abstract score, the judge's task is simply to count how many of these
assertions are satisfied. A percentage score can be derived from this
number, divided by the total number of assertions.
* **repetitions** - Rather than running each example once, and judging
it N times, we'll **run** the example N times. Right now, I'm just
judging the output once per run, because I believe that with these more
clear scoring criteria, the main source of non-determinism will be the
*agent's* behavior, not the judge's
### Questions
* **accounting for diagnostic errors** - Previously, the judge was asked
to incorporate diagnostics into their abstract scores. Now that the
"score" is determined directly from the criteria, the diagnostic will
not be captured in the score. How should the diagnostics be accounted
for in the eval? One thought is - let's simply count and report the
number of errors remaining after the agent finishes, as a separate field
of the run (along with diff score and thread score). We could consider
normalizing it using the total lines of added code (like errors per 100
lines of code added) in order to give it some semblance of stability
between examples.
* **repetitions** - How many repetitions should we run on CI? Each
repetition takes significant time, but I think running more than one
repetition will make the scores significantly less volatile.
### Todo
* [x] Fix `--concurrency` implementation so that only N tasks are
spawned
* [x] Support `--repetitions` efficiently (re-using the same worktree)
* [x] Restructure judge prompts to count passing criteria, not compute
abstract score
* [x] Report total number of diagnostics in some way
* [x] Format output nicely
Release Notes:
- N/A *or* Added/Fixed/Improved ...
---------
Co-authored-by: Antonio Scandurra <me@as-cii.com>
These are only written when the diff has changed. `patch.diff` has been
removed as its redundant with `last.diff`.
It can be convenient to open `last.diff` and use undo/redo to navigate
its history.
Release Notes:
- N/A
Split `locator` out of DebugTaskDefinition to make it clearer when
location needs to happen.
Release Notes:
- N/A
---------
Co-authored-by: Anthony Eid <hello@anthonyeid.me>
Co-authored-by: Anthony <anthony@zed.dev>
Co-authored-by: Cole Miller <m@cole-miller.net>
This PR adds a new `PlatformKeyboardLayout` trait with two methods:
`id(&self) -> &str` and `name(&self) -> &str`. The `id()` method returns
a unique identifier for the keyboard layout, while `name()` provides a
human-readable name. This distinction is especially important on
Windows, where the `id` and `name` can be quite different. For example,
the French layout has an `id` of `0000040C`, which is not
human-readable, whereas the `name` would simply be `French`. Currently,
the existing `keyboard_layout()` method returns what's essentially the
same as `id()` in this new design.
This PR implements the `name()` method for both Windows and macOS. On
Linux, for now, `name()` still returns the same value as `id()`.
Release Notes:
- N/A
Now that we've established a proper eval in tree, this PR is reboots of
our agent loop back to a set of minimal tools and simpler prompts. We
should aim to get this branch feeling subjectively competitive with
what's on main and then merge it, and build from there.
Let's invest in our eval and use it to drive better performance of the
agent loop. How you can help: Pick an example, and then make the outcome
faster or better. It's fine to even use your own subjective judgment, as
our evaluation criteria likely need tuning as well at this point. Focus
on making the agent work better in your own subjective experience first.
Let's focus on simple/practical improvements to make this thing work
better, then determine how we can craft our judgment criteria to lock
those improvements in.
Release Notes:
- N/A
---------
Co-authored-by: Max <max@zed.dev>
Co-authored-by: Antonio <antonio@zed.dev>
Co-authored-by: Agus <agus@zed.dev>
Co-authored-by: Richard <richard@zed.dev>
Co-authored-by: Max Brunsfeld <maxbrunsfeld@gmail.com>
Co-authored-by: Antonio Scandurra <me@as-cii.com>
Co-authored-by: Michael Sloan <mgsloan@gmail.com>
For some reason `SIGTTIN` sometimes gets sent to the process group,
causing it to stop when run from a terminal. This solves that issue by
putting the shell in a new session + progress group.
This allows removal of a workaround of using `exit 0;` to restore
handling of ctrl-c after exit. In testing this appears to no longer be
necessary.
Closes#27716
Release Notes:
- Fixed Zed sometimes becoming a stopped background process when run
from a terminal.
