improvement(desktop): ship the background executor without a flag, close its QA gaps by waleedlatif1 · Pull Request #8782 · simstudioai/sim

waleedlatif1 · 2026-10-08T03:41:36Z

Summary

Ship the desktop background executor without the mothership-desktop-background-executor flag. The executor now runs wherever Sim can track desktop presence, which means wherever Redis is configured. Installs without Redis keep running desktop tools in the chat window. Organization surfaces never show desktop activity, and runs that are already bound are unaffected.
The flag gate becomes isDesktopBackgroundExecutorAvailable(). The client learns availability from a server-computed desktopExecutorAvailable prop on the workspace sidebar instead of the feature-flag map. I removed the registry entry, the MSHIP_DESKTOP_BACKGROUND_EXECUTOR fallback and the CI env export.
Fix a background terminal run that was refused when the shell's startup files stopped to wait for input, such as an update prompt. The app waited a flat 8 s for the first prompt and then blamed shell integration. The generated zsh/bash startup files now emit a startup marker before the user's rc runs:
- With no marker within 8 s, the run is refused as before.
- With the marker, the shell gets 30 s from the marker to reach a prompt. Past that, the run is refused with the screen the shell is stuck on.
- Stop ends the wait.
Close the executor's QA gaps:
- E2E reports are written per check, so a Playwright worker restart can no longer wipe a failure.
- The fixture shells run with an empty ZDOTDIR.
- The fixture Sim cuts the network at the socket level.
- A new test shows that a call declined in a background chat is never claimed or run.
Update the live desktop suite for the executor being on. Its Sim has Redis, so every turn now binds to the app.
- The chat-view tests run against a registration answered the way an install without Redis answers it.
- Two new tests let Sim's real answer through: a call issued after the user switches chats runs on the desktop, and a result reported across a network cut reaches the agent exactly once. In that test the first report lands late and the retry is deduplicated.

After deploy: the mothership-desktop-background-executor entry in the staging AppConfig feature-flags document is unused. An operator should remove it. The infra validator only checks the document's shape, so it needs no change.

Type of Change

Feature
Bug fix

Testing

Type checks for apps/sim and apps/desktop pass, as do biome, check:comment-hygiene, check:api-validation, check:client-boundary, check:application-graph and check:unused-exports.
Sim unit tests pass: 49 files, 381 tests across the sidebar, workspace and org layouts, lib/desktop and the feature-flag config. A new layout test checks that the sidebar receives presence availability.
Integration tests for the executor, bound turns and activity pass, 51 of 51, with no flag mock. Each case where presence is unavailable is checked to write nothing and leave turns unbound.
Red runs: the layout tests fail on the old layout, and the presence-unavailable integration cases fail when the gate is forced on.
Desktop vitest: 1068 tests pass. These include a regression test for a startup marker that arrives late, which checks that the 30 s bound runs from the marker.
Playwright, run locally:
- background-executor, terminal-cancel and check-report: 25 of 25 pass.
- desktop-tools-live-sim against a local Sim with Redis: 12 of 12 pass. Before the suite update, it failed from the first test.

Checklist

Code follows project style guidelines
Self-reviewed my changes
Tests added/updated and passing
No new warnings introduced
I confirm that I have read and agree to the terms outlined in the Contributor License Agreement (CLA)

…a run A background terminal run on a fresh shell waited a flat 8 s for the first prompt and then failed with NO_SHELL_INTEGRATION. A startup file that stops at a question (oh-my-zsh's update prompt) holds the prompt indefinitely, so the run failed with a message blaming the shell's integration. The generated startup files now send a SimStartup marker before any of the user's files run. A shell that never sends it within 8 s of spawning is not instrumented and is refused as before (an unsupported shell at once). One that did gets 30 s to reach its prompt, which tolerates slow startup files on a busy machine; past that the run is refused with the screen it is stuck on, so the agent can ask the user to answer it. Stop now ends the wait, and a shell that exits mid-wait reports SESSION_CLOSED.

…rk cut - background-executor and terminal-cancel record each check through one shared helper as it finishes. The background-executor report was written from module state in afterAll, so a worker restart after a failure wiped it and the report could read green. Reports from an earlier run are replaced, not appended to. check-report.spec.ts forces a worker restart in a nested Playwright run and asserts the failure stays. - The executor fixture launches the app with zsh and an empty ZDOTDIR, so a developer's .zshrc (an unanswered oh-my-zsh update prompt) cannot hold the first prompt and fail unrelated scenarios. - Scenario B now cuts the network at the socket level: every open connection, the doorbell stream included, drops and new ones are reset. It asserts the result lands exactly once, the doorbell reopens, and new work is picked up after reconnecting.

…or run The fixture Sim can now decline a held call the way Sim settles it, and the executor suite checks the device never claims it and its command never runs after the approval notification.

…r on The live suite's Sim has Redis, so with the executor shipped always-on every turn now binds to the app and runs in the background. The proxy answers the app's registration as an install without Redis would, so the chat-view tests keep covering that path, and the round trip asserts the app stays dormant. Two tests let Sim's own answer through: a call issued after the user switched chats runs on the desktop, and a result reported across a network cut (every connection dropped, the first report landing late) reaches the agent exactly once.

vercel · 2026-10-08T03:41:39Z

The latest updates on your projects. Learn more about Vercel for GitHub.

1 Skipped Deployment

Project	Deployment	Actions	Updated
docs	Skipped		Oct 8, 2026 4:15am UTC

waleedlatif1 · 2026-10-08T03:41:40Z

@greptile

waleedlatif1 · 2026-10-08T03:41:41Z

@cubic-dev-ai review this PR

cubic-dev-ai · 2026-10-08T03:41:46Z

@cubic-dev-ai review this PR

@waleedlatif1 I have started the AI code review. It will take a few minutes to complete.

cubic-dev-ai

All reported issues were addressed across 38 files

_{Reply with feedback, questions, or to request a fix.

Turn on auto-fix | Re-trigger cubic}

cubic-dev-ai

All reported issues were addressed across 38 files

_{Reply with feedback, questions, or to request a fix.

Turn on auto-fix | Re-trigger cubic}

…iles began The 30 s bound ran from spawn, so a shell whose startup marker arrived late on a busy machine got less than the promised time for its startup files. The deadline now starts at the marker.

greptile-apps · 2026-10-08T03:51:18Z

Confidence Score: 5/5

[Medium risk] Removes a feature flag for the background executor.

The PR appears safe to merge, with the previous helper finding fixed and no new actionable issues.

What we checked:

Startup waits still end safely: startupBegunAt is set only once. The wait has a fixed deadline, and Stop ends it before the command runs.

Summary

The PR enables the desktop background executor wherever Redis supports presence tracking, improves shell-startup waits, and strengthens end-to-end checks.

Since the last review, shells receive their full startup allowance from the first startup marker.
The previous helper finding is fixed: check-report.spec.ts now uses omit, as waleedlatif1 reported.
No new actionable issues or mounted-rule violations were found.

Diagram

%%{init: {'theme': 'neutral'}}%%
flowchart TD
  Spawn[Shell starts] --> Marker{Startup marker within 8 seconds?}
  Marker -->|No| Refuse[Refuse command]
  Marker -->|Yes| Wait[Wait up to 30 seconds from marker]
  Wait --> Prompt{Prompt arrives?}
  Prompt -->|Yes| Run[Run command]
  Prompt -->|No| Screen[Refuse and show startup screen]
  Wait -->|Stop| Cancel[Cancel without running]

_{Reviews (2) · Last reviewed commit: "test(desktop): drop the outer runner's v..." · Reviewed by Greptile}

… helper

waleedlatif1 · 2026-10-08T04:15:20Z

@greptile

waleedlatif1 · 2026-10-08T04:15:21Z

@cubic-dev-ai review this PR

cubic-dev-ai · 2026-10-08T04:15:27Z

@cubic-dev-ai review this PR

@waleedlatif1 I have started the AI code review. It will take a few minutes to complete.

cubic-dev-ai

No issues found across 38 files

Confidence score: 5/5

Automated review surfaced no issues in the provided summaries.
No files require special attention.

_{Turn on auto-fix | Re-trigger cubic}

waleedlatif1 added 5 commits October 7, 2026 19:40

improvement(desktop): ship the background executor without a flag

431dd2e

test(desktop): a call declined in a background chat is never claimed …

5b6ce1b

…or run The fixture Sim can now decline a held call the way Sim settles it, and the executor suite checks the device never claims it and its command never runs after the approval notification.

cubic-dev-ai Bot reviewed Oct 8, 2026

View reviewed changes

Comment thread apps/desktop/src/main/terminal/session.ts Outdated

cubic-dev-ai Bot reviewed Oct 8, 2026

View reviewed changes

Comment thread apps/desktop/src/main/terminal/session.ts Outdated

Comment thread apps/sim/lib/api/contracts/desktop-executor.ts Outdated

vercel Bot deployed to Preview October 8, 2026 03:46 View deployment

fix(desktop): measure a shell's startup bound from when its startup f…

b39a245

…iles began The 30 s bound ran from spawn, so a shell whose startup marker arrived late on a busy machine got less than the promised time for its startup files. The deadline now starts at the marker.

waleedlatif1 mentioned this pull request Oct 8, 2026

v0.9.15: mship improvements, memory improvements, nextjs bump, plane and ramp integrations, outbox hardening #8748

Merged

greptile-apps Bot reviewed Oct 8, 2026

View reviewed changes

Comment thread apps/desktop/e2e/check-report.spec.ts Outdated

test(desktop): drop the outer runner's variables with the shared omit…

405f729

… helper

vercel Bot temporarily deployed to Preview October 8, 2026 04:15 Inactive

cubic-dev-ai Bot reviewed Oct 8, 2026

View reviewed changes

waleedlatif1 merged commit 31a911d into staging Oct 8, 2026
41 checks passed

waleedlatif1 deleted the improvement/desktop-executor-always-on branch October 8, 2026 04:20

waleedlatif1 mentioned this pull request Oct 8, 2026

fix(desktop): poll desktop activity only for users with a desktop, keep XDG zsh configs integrated #8785

Merged

6 tasks done

Conversation

waleedlatif1 commented Oct 8, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Type of Change

Testing

Checklist

Uh oh!

vercel Bot commented Oct 8, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

waleedlatif1 commented Oct 8, 2026

Uh oh!

waleedlatif1 commented Oct 8, 2026

Uh oh!

cubic-dev-ai Bot commented Oct 8, 2026

Uh oh!

cubic-dev-ai Bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

cubic-dev-ai Bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

greptile-apps Bot commented Oct 8, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Confidence Score: 5/5

Summary

Uh oh!

Uh oh!

waleedlatif1 commented Oct 8, 2026

Uh oh!

waleedlatif1 commented Oct 8, 2026

Uh oh!

cubic-dev-ai Bot commented Oct 8, 2026

Uh oh!

cubic-dev-ai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

This branch was previously deployed

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

waleedlatif1 commented Oct 8, 2026 •

edited

Loading

vercel Bot commented Oct 8, 2026 •

edited

Loading

cubic-dev-ai Bot left a comment •

edited

Loading

cubic-dev-ai Bot left a comment •

edited

Loading

greptile-apps Bot commented Oct 8, 2026 •

edited

Loading