Add ./script/format for customized cargo fmt invocation. (warpdotdev#11747)

## Description

This adds a formatting script at `./script/format` that invokes `cargo
fmt` with the options we want for our codebase.

Due to the fact that we're using some unstable formatter options, we
need to pass `RUSTC_BOOTSTRAP=1` to enable use of unstable options with
a stable toolchain, and that requires us to put together a small script
like this.

This PR updates existing formatting using the new script, and modifies
both `./script/presubmit` ~and CI to enforce formatting against this new
configuration~.

EDIT: Realized we probably want `./script/format` to soak for a week or
so before we start enforcement, due to us not using a merge queue.
Otherwise, we'll probably end up with non-conforming formatting changes
being merged into `master` and causing issues.

Author: vorporeal

.agents/skills/promote-feature/SKILL.md

@@ -81,7 +81,7 @@ pub const PREVIEW_FLAGS: &[FeatureFlag] = &[
 ### Validate
 
 ```bash
-cargo fmt
+./script/format
 cargo clippy --workspace --all-targets --all-features --tests -- -D warnings
 ```
 

.agents/skills/remove-feature-flag/SKILL.md

@@ -116,7 +116,7 @@ After removing the flag:
 
 ```bash
 # Format and lint
-cargo fmt
+./script/format
 cargo clippy --workspace --all-targets --all-features --tests -- -D warnings
 
 # Run tests

.agents/skills/rust-unit-tests/SKILL.md

@@ -106,7 +106,7 @@ cargo test --doc
 ## Linting and formatting
 Run before submitting changes:
 ```bash
-cargo fmt
+./script/format
 cargo clippy --workspace --all-targets --all-features --tests -- -D warnings
 ```
 

.github/workflows/ci.yml

@@ -613,8 +613,9 @@ jobs:
         run:
           cargo metadata --locked --format-version=1 > ${{ matrix.null_device }} || (echo "::error::Cargo.lock is out-of-date with Cargo.toml. Run 'cargo check' to update." && exit 1)
 
-      - name: Run cargo fmt
-        run: cargo fmt --check
+      - name: Run ./script/format
+        shell: bash
+        run: ./script/format --check
 
       - name: Run cargo clippy
         shell: bash
@@ -724,8 +725,15 @@ jobs:
         run:
           cargo metadata --locked --format-version=1 >/dev/null || (echo "::error::Cargo.lock is out-of-date with Cargo.toml. Run 'cargo check' to update." && exit 1)
 
-      - name: Run cargo fmt
-        run: cargo fmt --check
+      - name: Check Rust formatting
+        shell: bash
+        run: |
+          # TODO(vorporeal): Once people have gotten used to ./script/format, in a week or so,
+          # we can have CI enforce the new formatting.
+          #./script/format --check
+
+          # Until then, we'll keep running the traditional formatting check.
+          cargo fmt --check
 
       - name: Run cargo clippy
         run: |

CONTRIBUTING.md

@@ -171,7 +171,7 @@ Run unit tests with `cargo nextest run`.
 
 ## Code Style
 
-- `cargo fmt` and `cargo clippy --workspace --all-targets --all-features --tests -- -D warnings` must pass.
+- `./script/format --check` and `cargo clippy --workspace --all-targets --all-features --tests -- -D warnings` must pass.
 - Prefer imports over path qualifiers, inline format args (`println!("{x}")`), and exhaustive `match` over `_` wildcards.
 - See [WARP.md](WARP.md) for the full style guide, including WarpUI patterns and terminal model locking rules.
 

FAQ.md

@@ -80,7 +80,7 @@ No. Contributing by hand or with your own agent is free. Oz runs on Warp's credi
 
 ### Are agent-generated PRs held to the same bar as human ones?
 
-Yes. The same Oz + SME review, the same tests, and the same `cargo fmt` / `cargo clippy` / presubmit checks apply regardless of who (or what) wrote the code. Whether a PR is hand-written or agent-written doesn't change the quality bar — it changes how quickly you can iterate to meet it.
+Yes. The same Oz + SME review, the same tests, and the same `./script/format` / `cargo clippy` / presubmit checks apply regardless of who (or what) wrote the code. Whether a PR is hand-written or agent-written doesn't change the quality bar — it changes how quickly you can iterate to meet it.
 
 ### Will my issues, comments, or code be used to train models?
 

WARP.md

@@ -31,7 +31,7 @@ Environment variables:
 
 ### Linting and Formatting
 - `./script/presubmit` - Run all presubmit checks (fmt, clippy, tests)
-- `cargo fmt` - Format code
+- `./script/format` - Format code
 - `cargo clippy --workspace --all-targets --all-features --tests -- -D warnings` - Run clippy
 - `./script/run-clang-format.py -r --extensions 'c,h,cpp,m' ./crates/warpui/src/ ./app/src/` - Format C/C++/Obj-C code
 - `find . -name "*.wgsl" -exec wgslfmt --check {} +` - Check WGSL shader formatting
@@ -129,9 +129,9 @@ This is a Rust-based terminal emulator with a custom UI framework called **WarpU
   ```
 
 **Pull Request Workflow**:
-- **ALWAYS** run cargo fmt and cargo clippy (the versions specified in ./script/presubmit) before opening a PR or pushing updates to an existing PR branch
+- **ALWAYS** run `./script/format` and `cargo clippy` (the versions specified in ./script/presubmit) before opening a PR or pushing updates to an existing PR branch
 - Those commands must pass completely before creating or updating a pull request
-- Specifically, ensure `cargo fmt` and `cargo clippy` checks pass
+- Specifically, ensure `./script/format` and `cargo clippy` checks pass
 - If they fail, fix all issues before proceeding with the PR
 - This applies to:
   - Opening new pull requests

app/src/ai/agent/api/convert_from.rs

@@ -2,8 +2,7 @@
 use std::collections::HashMap;
 use std::time::Duration;
 
-use ai::agent::action::LifecycleEventType as StartAgentLifecycleEventType;
-use ai::agent::action::ReadSkillRequest;
+use ai::agent::action::{LifecycleEventType as StartAgentLifecycleEventType, ReadSkillRequest};
 use ai::agent::action_result::StartAgentVersion;
 use ai::agent::convert::ToolToAIAgentActionError;
 use ai::agent::UnknownCitationTypeError;

app/src/ai/agent/api/convert_from_tests.rs

@@ -1,6 +1,7 @@
+use std::path::PathBuf;
+
 use ai::agent::action::AskUserQuestionType;
 use ai::skills::{SkillPathOrigin, SkillReference};
-use std::path::PathBuf;
 use warp_multi_agent_api as api;
 use warp_util::local_or_remote_path::LocalOrRemotePath;
 

app/src/ai/agent/conversation_tests.rs

@@ -1,6 +1,6 @@
-use ai::api_keys::ApiKeyManager;
 use std::collections::HashMap;
 
+use ai::api_keys::ApiKeyManager;
 use warp_core::features::FeatureFlag;
 use warp_multi_agent_api as api;
 use warpui::{App, SingletonEntity};
@@ -11,7 +11,8 @@ use super::{
 };
 use crate::ai::artifacts::Artifact;
 use crate::ai::llms::LLMPreferences;
-use crate::auth::{auth_manager::AuthManager, AuthStateProvider};
+use crate::auth::auth_manager::AuthManager;
+use crate::auth::AuthStateProvider;
 use crate::network::NetworkStatus;
 use crate::persistence::model::AgentConversationData;
 use crate::server::server_api::ServerApiProvider;