Preserve background terminals on interrupt and rename cleanup command to /stop by friel-openai · Pull Request #14602 · openai/codex

friel-openai · 2026-03-13T15:38:45Z

Motivation

Interrupting a running turn (Ctrl+C / Esc) currently also terminates long‑running background shells, which is surprising for workflows like local dev servers or file watchers.
The existing cleanup command name was confusing; callers expect an explicit command to stop background terminals rather than a UI clear action.
Make background‑shell termination explicit and surface a clearer command name while preserving backward compatibility.

Description

Renamed the background‑terminal cleanup slash command from Clean (/clean) to Stop (/stop) and kept clean as an alias in the command parsing/visibility layer, updated the user descriptions and command popup wiring accordingly.
Updated the unified‑exec footer text and snapshots to point to /stop (and trimmed corresponding snapshot output to match the new label).
Changed interrupt behavior so Op::Interrupt (Ctrl+C / Esc interrupt) no longer closes or clears tracked unified exec / background terminal processes in the TUI or core cleanup path; background shells are now preserved after an interrupt.
Updated protocol/docs to clarify that turn/interrupt (or Op::Interrupt) interrupts the active turn but does not terminate background terminals, and that thread/backgroundTerminals/clean is the explicit API to stop those shells.
Updated unit/integration tests and insta snapshots in the TUI and core unified‑exec suites to reflect the new semantics and command name.

Testing

Ran formatting with just fmt in codex-rs (succeeded).
Ran cargo test -p codex-protocol (succeeded).
Attempted cargo test -p codex-tui but the build could not complete in this environment due to a native build dependency that requires libcap development headers (the codex-linux-sandbox vendored build step); install libcap-dev / make libcap.pc available in PKG_CONFIG_PATH to run the TUI test suite locally.
Updated and accepted the affected insta snapshots for the TUI changes so visual diffs reflect the new /stop wording and preserved interrupt behavior.

Codex Task

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: a4dac1c67d

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

chatgpt-codex-connector · 2026-03-13T15:43:38Z

codex-rs/tui/src/slash_command.rs

    Rollout,
    Ps,
-    Clean,
+    #[strum(serialize = "stop", serialize = "clean")]


Make /stop the canonical serialized name for Stop command

Defining Stop as #[strum(serialize = "stop", serialize = "clean")] breaks lookup: command() uses IntoStaticStr, which selects Strum's preferred serialization (longest string), so this variant resolves to "clean". find_builtin_command matches only that canonical name, so /stop is not recognized even though UI/docs now instruct users to run /stop.

Useful? React with 👍 / 👎.

@codex fix comments

Summary

Fixed slash-command canonicalization for Stop by removing IntoStaticStr usage and making SlashCommand::command() explicitly return "stop" for SlashCommand::Stop while keeping existing behavior (as_ref()) for other commands. This ensures /stop is the canonical emitted name used by popup/lookup paths. clean remains as a parse alias via the existing Strum serialize attributes on the enum variant. codex-rs/tui/src/slash_command.rsL1-L8 codex-rs/tui/src/slash_command.rsL48-L49 codex-rs/tui/src/slash_command.rsL113-L120

Added a regression unit test to lock in that SlashCommand::Stop.command() is "stop". codex-rs/tui/src/slash_command.rsL200-L208

Testing

✅ just fmt

⚠️ cargo test -p codex-tui (fails in this environment because codex-linux-sandbox’s build script cannot find libcap/libcap.pc via pkg-config)

⚠️ curl -I https://docs.rs/strum/latest/strum/derive.IntoStaticStr.html | head (web lookup attempt blocked by proxy with HTTP 403)

Committed as 664b77a and created a follow-up PR via the make_pr tool.

View task →

@codex No, I don't like that design. We shouldn't change the derive in codex-rs/tui/src/slash_command.rs, instead we should fix the enum variant or member.

I'm skeptical of your design.

Isn't it simpler to

#[strum(serialize = "stop", serialize = "clean")] Stop,

Change the command variant to Clean and allow it to accept stop as an alternative serialization?

Summary

Updated the Stop slash-command variant metadata to use Strum’s preferred-name override (to_string = "stop") while keeping "clean" as a parse alias, so canonical command emission remains /stop without changing enum derives or command() behavior. codex-rs/tui/src/slash_command.rsL51-L52 codex-rs/tui/src/slash_command.rsL116-L120

Added regression tests to lock in both behaviors: canonical name for Stop is "stop", and "clean" still parses to SlashCommand::Stop. codex-rs/tui/src/slash_command.rsL200-L215

Committed the change as 0490d8b and created a follow-up PR via the make_pr tool.

Testing

✅ just fmt

⚠️ cargo test -p codex-tui (fails in this environment because codex-linux-sandbox requires system libcap/libcap.pc via pkg-config)

View task →

@codex fix these tests. the first might be a red herring - a race? but look into it anyhow

────────────
Nextest run ID 5b5b8d29-401d-4c11-abc6-3ee4f5e0d6e5 with nextest profile: default
Starting 5612 tests across 132 binaries (28 tests skipped)
SLOW [> 15.000s] (─────────) codex-core unified_exec::tests::unified_exec_timeouts
FAIL [ 0.454s] (3244/5612) codex-core::all suite::unified_exec::unified_exec_interrupt_preserves_long_running_session
stdout ───

running 1 test test suite::unified_exec::unified_exec_interrupt_preserves_long_running_session ... FAILED failures: failures: suite::unified_exec::unified_exec_interrupt_preserves_long_running_session test result: FAILED. 0 passed; 1 failed; 0 ignored; 0 measured; 803 filtered out; finished in 0.44s

stderr ───
kill: 1155: No such process

thread 'suite::unified_exec::unified_exec_interrupt_preserves_long_running_session' (257029) panicked at core/tests/suite/unified_exec.rs:2096:5: expected unified exec process to remain alive after interrupt stack backtrace: 0: __rustc::rust_begin_unwind at /rustc/254b59607d4417e9dffbc307138ae5c86280fe4c/library/std/src/panicking.rs:689:5 1: core::panicking::panic_fmt at /rustc/254b59607d4417e9dffbc307138ae5c86280fe4c/library/core/src/panicking.rs:80:14 2: all::suite::unified_exec::unified_exec_interrupt_preserves_long_running_session::{{closure}} at ./tests/suite/unified_exec.rs:2096:5 3: <core::pin::Pin<P> as core::future::future::Future>::poll at /Users/runner/.rustup/toolchains/1.93.0-aarch64-apple-darwin/lib/rustlib/src/rust/library/core/src/future/future.rs:133:9 4: tokio::runtime::park::CachedParkThread::block_on::{{closure}} at /Users/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/tokio-1.49.0/src/runtime/park.rs:284:71 5: tokio::task::coop::with_budget at /Users/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/tokio-1.49.0/src/task/coop/mod.rs:167:5 6: tokio::task::coop::budget at /Users/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/tokio-1.49.0/src/task/coop/mod.rs:133:5 7: tokio::runtime::park::CachedParkThread::block_on at /Users/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/tokio-1.49.0/src/runtime/park.rs:284:31 8: tokio::runtime::context::blocking::BlockingRegionGuard::block_on at /Users/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/tokio-1.49.0/src/runtime/context/blocking.rs:66:14 9: tokio::runtime::scheduler::multi_thread::MultiThread::block_on::{{closure}} at /Users/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/tokio-1.49.0/src/runtime/scheduler/multi_thread/mod.rs:89:22 10: tokio::runtime::context::runtime::enter_runtime at /Users/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/tokio-1.49.0/src/runtime/context/runtime.rs:65:16 11: tokio::runtime::scheduler::multi_thread::MultiThread::block_on at /Users/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/tokio-1.49.0/src/runtime/scheduler/multi_thread/mod.rs:88:9 12: tokio::runtime::runtime::Runtime::block_on_inner at /Users/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/tokio-1.49.0/src/runtime/runtime.rs:370:50 13: tokio::runtime::runtime::Runtime::block_on at /Users/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/tokio-1.49.0/src/runtime/runtime.rs:342:18 14: all::suite::unified_exec::unified_exec_interrupt_preserves_long_running_session at ./tests/suite/unified_exec.rs:2104:7 15: all::suite::unified_exec::unified_exec_interrupt_preserves_long_running_session::{{closure}} at ./tests/suite/unified_exec.rs:2021:69 16: core::ops::function::FnOnce::call_once at /Users/runner/.rustup/toolchains/1.93.0-aarch64-apple-darwin/lib/rustlib/src/rust/library/core/src/ops/function.rs:250:5 17: core::ops::function::FnOnce::call_once at /rustc/254b59607d4417e9dffbc307138ae5c86280fe4c/library/core/src/ops/function.rs:250:5 note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace. FAIL [ 0.221s] (4582/5612) codex-tui bottom_pane::slash_commands::tests::clean_command_alias_resolves_for_dispatch

stdout ───

running 1 test test bottom_pane::slash_commands::tests::clean_command_alias_resolves_for_dispatch ... FAILED failures: failures: bottom_pane::slash_commands::tests::clean_command_alias_resolves_for_dispatch test result: FAILED. 0 passed; 1 failed; 0 ignored; 0 measured; 1316 filtered out; finished in 0.21s

stderr ───

thread 'bottom_pane::slash_commands::tests::clean_command_alias_resolves_for_dispatch' (273078) panicked at tui/src/bottom_pane/slash_commands.rs:95:9: assertion failed: `(left == right)` Diff < left / right > : <None >Some( > Stop, >) stack backtrace: 0: __rustc::rust_begin_unwind at /rustc/254b59607d4417e9dffbc307138ae5c86280fe4c/library/std/src/panicking.rs:689:5 1: core::panicking::panic_fmt at /rustc/254b59607d4417e9dffbc307138ae5c86280fe4c/library/core/src/panicking.rs:80:14 2: codex_tui::bottom_pane::slash_commands::tests::clean_command_alias_resolves_for_dispatch at ./src/bottom_pane/slash_commands.rs:95:9 3: codex_tui::bottom_pane::slash_commands::tests::clean_command_alias_resolves_for_dispatch::{{closure}} at ./src/bottom_pane/slash_commands.rs:94:51 4: core::ops::function::FnOnce::call_once at /Users/runner/.rustup/toolchains/1.93.0-aarch64-apple-darwin/lib/rustlib/src/rust/library/core/src/ops/function.rs:250:5 5: core::ops::function::FnOnce::call_once at /rustc/254b59607d4417e9dffbc307138ae5c86280fe4c/library/core/src/ops/function.rs:250:5 note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.

────────────
Summary [ 165.095s] 5612 tests run: 5610 passed (1 slow), 2 failed, 28 skipped
FAIL [ 0.454s] (3244/5612) codex-core::all suite::unified_exec::unified_exec_interrupt_preserves_long_running_session
FAIL [ 0.221s] (4582/5612) codex-tui bottom_pane::slash_commands::tests::clean_command_alias_resolves_for_dispatch
error: test run failed

Summary

Updated built-in slash-command resolution to support both canonical names and aliases in dispatch: find_builtin_command now first matches the visible canonical command string, then falls back to parsing aliases (like clean) and maps them back through visible built-ins so /clean correctly resolves to SlashCommand::Stop. codex-rs/tui/src/bottom_pane/slash_commands.rsL39-L55

Changed unified-exec process lifecycle so spawned processes are persisted before waiting for the initial yield window; this prevents interrupt-time cancellation from dropping an in-flight long-running process before it is tracked as a background terminal. Also simplified post-collection cleanup so exited processes are released from the store in one path. codex-rs/core/src/unified_exec/process_manager.rsL153-L267

Testing

✅ just fmt

⚠️ cargo test -p codex-tui clean_command_alias_resolves_for_dispatch -- --nocapture (build blocked in this environment: missing system libcap / libcap.pc required by codex-linux-sandbox)

⚠️ cargo test -p codex-core unified_exec_interrupt_preserves_long_running_session -- --nocapture (same libcap environment limitation)

Committed as 1f8ef85, and I also created the follow-up PR via the make_pr tool.

View task →

Preserve background unified-exec processes across interrupts, keep /stop as the cleanup command, and collapse the branch back to a single commit for easier future restacks.

friel-openai added the codex Label used by connector to tag PRs that have been reviewed by Codex label Mar 13, 2026 — with ChatGPT Codex Connector

chatgpt-codex-connector bot reviewed Mar 13, 2026

View reviewed changes

friel-openai force-pushed the dev/codex/rename-/clear-and-modify-ctrl-c-behavior branch 2 times, most recently from 2e91351 to 8334001 Compare March 14, 2026 20:29

fix(unified-exec): preserve background terminals on interrupt

2560335

Preserve background unified-exec processes across interrupts, keep /stop as the cleanup command, and collapse the branch back to a single commit for easier future restacks.

friel-openai force-pushed the dev/codex/rename-/clear-and-modify-ctrl-c-behavior branch from 8334001 to 2560335 Compare March 14, 2026 20:32

github-actions bot mentioned this pull request Mar 15, 2026

📊 AI CLI 工具社区动态日报 2026-03-15 gsscsd/big_model_radar#37

Open

daveaitel-openai approved these changes Mar 15, 2026

View reviewed changes

This was referenced Mar 16, 2026

📊 AI CLI 工具社区动态日报 2026-03-16 gsscsd/big_model_radar#43

Open

📊 Bản tin hàng ngày công cụ AI CLI 2026-03-16 compasify/agents-radar#50

Open

friel-openai merged commit ba463a9 into main Mar 16, 2026
32 checks passed

friel-openai deleted the dev/codex/rename-/clear-and-modify-ctrl-c-behavior branch March 16, 2026 05:17

github-actions bot locked and limited conversation to collaborators Mar 16, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Preserve background terminals on interrupt and rename cleanup command to /stop#14602

Preserve background terminals on interrupt and rename cleanup command to /stop#14602
friel-openai merged 1 commit intomainfrom
dev/codex/rename-/clear-and-modify-ctrl-c-behavior

friel-openai commented Mar 13, 2026

Uh oh!

chatgpt-codex-connector bot left a comment

Uh oh!

chatgpt-codex-connector bot Mar 13, 2026

Uh oh!

friel-openai Mar 13, 2026

Uh oh!

chatgpt-codex-connector bot Mar 13, 2026

Uh oh!

friel-openai Mar 13, 2026

Uh oh!

chatgpt-codex-connector bot Mar 13, 2026

Uh oh!

friel-openai Mar 14, 2026

Uh oh!

chatgpt-codex-connector bot Mar 14, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

friel-openai commented Mar 13, 2026

Motivation

Description

Testing

Uh oh!

chatgpt-codex-connector bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

chatgpt-codex-connector bot Mar 13, 2026

Choose a reason for hiding this comment

Uh oh!

friel-openai Mar 13, 2026

Choose a reason for hiding this comment

Uh oh!

chatgpt-codex-connector bot Mar 13, 2026

Choose a reason for hiding this comment

Summary

Testing

Uh oh!

friel-openai Mar 13, 2026

Choose a reason for hiding this comment

Uh oh!

chatgpt-codex-connector bot Mar 13, 2026

Choose a reason for hiding this comment

Uh oh!

friel-openai Mar 14, 2026

Choose a reason for hiding this comment

Uh oh!

chatgpt-codex-connector bot Mar 14, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants