Skip to content

fix: stop calling a failed cloud launch a success - #34844

Merged
KiroCrewCaptain merged 1 commit into
mainfrom
fix/34713-failed-launch-status
Nov 13, 2026
Merged

KiroCrewCaptain merged 1 commit into
mainfrom
fix/34713-failed-launch-status

Conversation

@JLargent42

@JLargent42 JLargent42 commented Nov 10, 2026 •

Copy link
Copy Markdown
Contributor

Problem / Motivation

Goal: After a cloud launch fails, the line next to the Launch buttons says it failed and offers a retry the user can change, instead of calling it a success.

In Settings → Remote Crew → Set up a new one, after a launch the progress card below shows "Launch failed Kiro Crew Cloud (kc-…)", yet the status line beside the buttons still reads "Launched. Its progress is shown below. To start a second crew with the same settings, choose Launch another crew." Launch stays greyed out, and the only offered action repeats the settings that just failed.

Why it matters

Anyone whose first cloud launch fails (a subnet with no egress, a missing permission) is told it worked and steered to a repeat that fails the same way. The onboarding bug bash hit this in 2 runs. It is the first thing a new user reads after a failure.

Not a goal

  • Changing the failure message in the progress card, or what the gateway reports.
  • Cancelled launches: the user chose that, and the existing same-settings guard still applies to them.
  • Registered (edition) provisioner forms, which own their own copy and buttons.

What changed (motivation → approach → change)

Symptom: success copy and a same-settings "Launch another crew" after a failed launch. Root cause: the same-settings guard (alreadyLaunched) compares only the launch body key with the last one sent, and never looks at how that launch ended. So a failed launch counts as a crew that exists.

Change in RemoteCrewPanel.tsx, for the launch these settings went out with (activeJob.id === activeLaunchId) when its job ends failed:

  • The status line is always an inline ErrorNotice, never the success wording. It has no hand-off, because the subnet and sign-in choice the next attempt sends live only in this form's state.
  • The same-settings guard is lifted only when the launch stopped before provisioning started. In that case the provision step is still pending or skipped, there is no instance_id, and nothing registered, so nothing exists in the account. Launch itself is then the retry, "Launch another crew" is not shown, and the line reads "Fix the problem shown below, or change the Subnet ID above, then choose Launch to try again." Launch stays enabled with unchanged settings because many causes are fixed outside the form, and the issue's step 3 is "do what it says, then launch again".
  • From provisioning on, a stack may still exist, so the guard stays. Launch is off for identical settings, and "Launch another crew" or a changed Subnet ID starts the next attempt. The job is a snapshot: a crew it registered can be deleted afterwards, and an instance that never registered has no row. So the line is decided by the live crew list at the moment it renders:
    • The job registered and its instance_id is in the live instances list: "The launch did not finish. Its crew is listed in Your crews, where you can fix or delete it." (launch_failed_crew_listed)
    • Anything else: "The launch did not finish. Check Your crews before launching again." (launch_failed_check_crews)

The three new keys are in every catalog, and en-XA is regenerated. The retry line names the Launch button by its label in each locale, since main now translates it.

Backwards compatibility

Compatible: only the built-in launch form's state after a failed job changes. No API, config or stored data is touched.

Removes nothing: the "Launch another crew" button is still offered after a running or finished launch, and after any failure from provisioning on. It is hidden only after a failure before provisioning, where Launch itself repeats the settings.

Tests

  • website/src/test/RemoteCrewPanelLaunchForm.test.tsx: new case "after a launch that failed before provisioning, says it failed and offers Launch as the retry". The job is failed at preflight with provision pending. The test checks that the retry line is an alert, that the success line and "Launch another crew" are not shown, that Launch is enabled, and that a retry with a changed Subnet ID sends that subnet.

  • Same file: new it.each "after a launch that failed during %s, says it failed but keeps the same-settings guard". The instance list returns the crew only after the launch goes out, as the gateway's would. It has four cases:

    • a sign-in failure with the registered crew in the live list, which gets the listed line;
    • a sign-in failure with the registered crew deleted since, which gets the neutral line;
    • an instance that provisioned but never registered (present in the list), which gets the neutral line;
    • a provision failure with no instance, which gets the neutral line.

    In each, its own line is an alert and the other line is absent, the success line and the retry line are absent, Launch is off, "Launch another crew" is offered, and a changed Subnet ID re-enables Launch.

  • All of these fail on main. The listed case fails on the previous head's panel code, since that code never says "listed".

  • The existing same-settings guard test still passes.

Proof: not provable -- frontend test (prove.py runs Python only)

Local: 58 related frontend test files (run_scoped_tests.py --surface frontend) pass, plus npm run typecheck, eslint on the touched files, npm run i18n:check, and test/test_cloud_sizes_ui_parity.py, test/test_remote_crew_doc.py, test/test_file_line_cap.py and test/test_brand_name_gate.py at -n 2.

Manual verification

I rendered the real RemoteCrewPanel through website/capture/remote-crew-update.html with intercepted API answers: one launch whose job comes back failed, either at the provision step or at the AWS check. I captured it on main and with the fix (below). I did not run a real AWS launch.

Screenshots / video

Before (main): success wording and a same-settings "Launch another crew" beside the failed card.

before, dark

After, failed at the AWS check (nothing created): an inline error beside Launch points at the message and the Subnet ID, and Launch is the retry.

after, failed before provisioning, dark

After, failed at sign-in with the crew in the live list: the line says it is listed in Your crews. The card offers its own sign-in recovery.

after, failed after the crew registered, dark

After, failed during provisioning with no instance: the neutral line says to check Your crews before launching again. "Launch another crew" is the explicit repeat.

after, failed during provisioning, dark

After, failed during provisioning, light theme

after, failed during provisioning, light

Related Issues

Fixes #34713

Sibling sweep

Swept: git grep -n "launched_same_settings|alreadyLaunched" -- website/src
Siblings: none -- the same-settings guard and its status line exist only in the built-in form of RemoteCrewPanel.tsx

Pattern harvest

Not generalizable: a one-off guard that keyed on the request sent and never read the job outcome; no second launch guard exists to share a rule with.

Checklist

  • At most two commits (one is the norm), with a Conventional Commits title (feat|fix|docs|style|refactor|perf|test|chore|ci|build|revert: ...)
  • Existing tests pass and new tests added for new functionality
  • Self-review completed; code follows project style guidelines
  • Documentation updated (if applicable)
  • No secrets, credentials, or internal references in the diff

Contribution License Agreement

Contributions are accepted under the project's Apache-2.0 license.

@JLargent42
JLargent42 requested review from a team and KiroCrewCaptain as code owners November 10, 2026 22:05
@JLargent42 JLargent42 self-assigned this Nov 10, 2026
@github-actions

github-actions Bot commented Nov 10, 2026 •

Copy link
Copy Markdown
Contributor

First Principles Review (Opus 5.5) — ✅ PASS

Premise-level review of db7f6e2f4c0aa63a850681dbed735c827666ef83 — why this exists and whether the shipped surface is the smallest honest version. Updated in place on each push. A BLOCK verdict blocks PR readiness; PASS does not.

First-Principles-Verdict: PASS

Fixes #34713 at the cause: the guard now checks how the launch ended. One extra wording branch goes beyond what the Goal needs.

Notes

  • Item 3 — oversized: the Goal is that the line "says it failed and offers a retry the user can change". The neutral line ("Check Your crews before launching again") is already accurate when the crew is listed, so the launchFailedCrewListed branch and its instances.some(i => i.ssm_target === activeJob.instance_id) lookup are not needed for the fix. Subtraction: delete launchFailedCrewListed and the launch_failed_crew_listed key in all 13 locales, and keep launch_failed_check_crews for every failure from provisioning on. Clears when: the mechanism is removed, or a human amends the Goal.

What this change ships

Inventory (5 items) — 4 justified

Intent: a user whose cloud launch failed is told it failed and can retry with changed settings. This is a FIX.

  1. After a failed launch, the line beside Launch is an inline error instead of the "Launched…" success text — justified
  2. After a failure before provisioning, Launch is re-enabled with the same settings and acts as the retry — justified
  3. Failures from provisioning on get two different lines, chosen by whether the crew is in the live crew list — oversized: the neutral line alone covers the Goal
  4. "Launch another crew" is hidden after a failure before provisioning — justified
  5. Three new strings in every locale catalog, with en-XA regenerated — justified

[FIRST-PRINCIPLES-REVIEWED] db7f6e2

Description read

[DESCRIPTION-READ] d48e3de40dbf3cc9366ccd7ae1605608540dfd18a205db2ebfaa139f100944cb

That sha256 names the PR description this verdict read. Recompute it with .github/scripts/pr-description-capture.sh; if it differs, the description changed after this verdict was formed and any finding drawn from the description is unproven.

@github-actions

github-actions Bot commented Nov 10, 2026 •

Copy link
Copy Markdown
Contributor

Design Review (Opus 5.5) — ✅ PASS

Design-level review of db7f6e2f4c0aa63a850681dbed735c827666ef83 — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS does not.

Design-Verdict: PASS

Nothing blocks. The failed launch now reads the job's outcome, the guard is lifted only when no stack can exist yet, and screenshots of this revision are attached.

[DESIGN-REVIEWED] db7f6e2

Description read

[DESCRIPTION-READ] 595800bbb7cbea2fc1e5d3c9effb2133e5859fe9bec850e54cd03a26fa0126a5

That sha256 names everything this verdict read: the PR description, plus the 1 evidence file(s) the workflow collected from it and handed this review. It covers that evidence because the media strip replaces every attachment URL with the same placeholder, so an edit swapping one attachment for another would leave the description's captured text identical while what this review judged changed. Recompute it with .github/scripts/pr-description-capture.sh over the same inputs; if it differs, the description or that evidence changed after this verdict was formed and any finding drawn from either is unproven.

@github-actions

github-actions Bot commented Nov 10, 2026 •

Copy link
Copy Markdown
Contributor

UX Review (Opus 5.5) — ✅ PASS

UX-level review of db7f6e2f4c0aa63a850681dbed735c827666ef83 — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS does not.

UX-Verdict: PASS

No blocker. The blind reader read the new retry line and the re-enabled Launch correctly ("try again … sure"), and every new failure state has a screenshot.

Notes

  • launch_failed_crew_listed / launch_failed_check_crews: on shots 3 and 4 the blind reader said "'Your crews' isn't a link, so I don't know where that list is". Make "Your crews" a jump link to that section on the page, so the next step sits next to the fact that needs it.
  • launch_failed_check_crews: on shot 4 the reader would not dare press "Launch another crew" because "I don't know if the failed one is still running (and billing)". The line should say what to look for, e.g. "The launch did not finish. If a crew from it appears in Your crews, delete it before launching again."

[UX-REVIEWED] db7f6e2

Description read

[DESCRIPTION-READ] e21c1672c0c6101b6d43e4f9e07135a016607a9de99224f8db2ba8a3e805cbb3

That sha256 names everything this verdict read: the PR description, plus the 8 evidence file(s) the workflow collected from it and handed this review. It covers that evidence because the media strip replaces every attachment URL with the same placeholder, so an edit swapping one attachment for another would leave the description's captured text identical while what this review judged changed. Recompute it with .github/scripts/pr-description-capture.sh over the same inputs; if it differs, the description or that evidence changed after this verdict was formed and any finding drawn from either is unproven.

@github-actions

github-actions Bot commented Nov 10, 2026 •

Copy link
Copy Markdown
Contributor

Opus 5.5 Review — ✅ no blocking findings

Reviewed db7f6e2f4c0aa63a850681dbed735c827666ef83 — this comment is updated in place on each push.

Review details

No findings.

[OPUS-REVIEWED] db7f6e2

Verdict parsed from the review's SHA-scoped output markers for commit db7f6e2f4c0aa63a850681dbed735c827666ef83.

False positive or not applicable? A repository writer can comment:
/ai-review override fable db7f6e2f4c0aa63a850681dbed735c827666ef83 false-positive: <reason, 20 to 500 characters>

@github-actions

github-actions Bot commented Nov 10, 2026 •

Copy link
Copy Markdown
Contributor

GPT 6.1 Review — ✅ no blocking findings

GPT 6.1 completed its review of db7f6e2f4c0aa63a850681dbed735c827666ef83 and found no blocking issues.

This comment is updated in place on each push.

Review details

No findings.
[GPT-REVIEWED] db7f6e2

False positive or not applicable? A repository writer can comment:
/ai-review override gpt db7f6e2f4c0aa63a850681dbed735c827666ef83 false-positive: <reason, 20 to 500 characters>

@JLargent42
JLargent42 force-pushed the fix/34713-failed-launch-status branch from 4ff4dee to 13e95ea Compare November 10, 2026 22:46
@JLargent42

Copy link
Copy Markdown
Contributor Author

AI review disposition for 4ff4deea7094d8364a213ad3a7ea355a53ac284b, fixed in 13e95ea9334ba7f12371de19bc6920d27a6618db:

  • Opus 5.5 (blocking, RemoteCrewPanel.tsx:945): fixed. A failed job lifts the same-settings guard only when no crew survived: !activeJob.instance_id && !isRegistered(activeJob). instance_id is set only once engine.provision returns, and a failed provision is rolled back. So a sign-in refusal after register, or any later-step failure, keeps the guard and "Launch another crew". The comment that said a failed launch made no crew is corrected. New test: "keeps the same-settings guard when a launch failed after its crew was created".
  • Design note (activeJob not tied to the launched job): fixed in the same line. The guard also requires activeJob.id === activeLaunchId, the id the accepted launch returned.
  • GPT 6.1 (blocking, errors-use-error-notice, :1678): fixed. The failure branch renders through <ErrorNotice variant="inline"> with a No hand-off comment. The subnet and sign-in choice the retry sends live only in this form's state, because the store copy is dropped when the gateway accepts a launch. This matches the progress card's own job.error notice. The test asserts the line is an alert.
  • UX note (says "failed" twice): taken. The copy now leads with the next step: "Fix the problem shown below, or change the Subnet ID above, then choose Launch to try again." It is updated in every catalog, and the screenshots are re-shot.

@github-actions github-actions Bot added the readiness: action required A blocking check or review needs attention label Nov 10, 2026
@JLargent42
JLargent42 force-pushed the fix/34713-failed-launch-status branch from 13e95ea to 438803b Compare November 11, 2026 00:56
@JLargent42

Copy link
Copy Markdown
Contributor Author

AI review disposition for 13e95ea9334ba7f12371de19bc6920d27a6618db, fixed in 438803be547098ea79506c68280ecd4c631f6dc0:

  • GPT 6.1 (blocking, RemoteCrewPanel.tsx:951, missing instance ID is not proof no crew is left): fixed by narrowing. This is the second finding with the same root cause as the earlier Opus one, which asked whether a failed launch is safe to repeat. So the approach changed instead of adding a third condition. The guard is now lifted only when the launch stopped before provisioning started, meaning the provision step is still pending or skipped, there is no instance_id, and nothing registered. In that case nothing can exist in the account.
  • Any failure from provisioning on keeps the guard. That covers a provision failure whose best-effort rollback may not have confirmed. Launch is off for identical settings, and "Launch another crew" is the explicit repeat. The inline error reads "Read the message below first: a crew may still exist from this attempt. To try again, change the Subnet ID above, or choose Launch another crew to repeat these settings."
  • No backend field. Adding a "rollback confirmed" field to the job would widen the launch API for a UI fix, so this PR doesn't do it.
  • New test: it.each "after a launch that failed during provisioning, with no instance returned, says it failed but keeps the same-settings guard". It fails on 13e95ea9 and passes now. The screenshots in the description show both states.

@JLargent42
JLargent42 force-pushed the fix/34713-failed-launch-status branch from 438803b to 98805b6 Compare November 11, 2026 02:06
@JLargent42

Copy link
Copy Markdown
Contributor Author

UX Review disposition for 438803be547098ea79506c68280ecd4c631f6dc0, fixed in 98805b6f28e6b8231c9ed4fe14ae1771125122b5:

  • Blocker (check-crews line hedges what the code knows, and contradicts the "stack was removed" card): fixed. The string is split in two.
    • When the job reports a created crew (instance_id, or registered), the line says plainly: "The crew from this attempt was created and still exists: it is listed in Your crews, where you can fix or delete it."
    • After a provision failure with no instance, it no longer guesses. It points at the card: "The message below says whether this attempt's stack was removed."
  • Note (why "Launch another crew" shows in one shot and not the other): taken. Both lines now add "Launch is off because these settings were just sent", and name the two ways forward.
  • Link to the crew: not added. "Your crews" is the tab beside "Set up a new one" on the same panel, so the line names it instead of adding a link.
  • New screenshot of the crew-exists state is in the description. The test it.each now asserts each case's own line, and that the other case's line is absent.

@KiroCrewCaptain

KiroCrewCaptain commented Nov 11, 2026 •

Copy link
Copy Markdown
Collaborator

Code change approver recommendation: Request changes

Why: the decider found 2 problems and estimates an 80% chance that this change introduces a defect, at or above the 80% level where it asks for changes.

The main fix works: after a failed launch the line is an inline error instead of the success wording, and Launch is the retry when the launch failed before provisioning. But the new 'crew listed in Your crews' line compares the registry id with the job's EC2 instance id. The registry id is built from the crew's name and the EC2 id is stored as ssm_target, so that line never shows in real use. The test fixture gives both ids the same value, which hides the bug. Frontend lint is also failing on this head, so the new tests never ran in CI. Requesting changes: match on ssm_target and fix the fixture.

Problems found:

  • The 'Its crew is listed in Your crews' line that the PR describes never shows on the real EC2 path, because the registry id never equals the EC2 instance id. Users always get the neutral line instead. Fix: match on i.ssm_target (or ssh_host). (website/src/pages/settings/RemoteCrewPanel.tsx, medium)
  • The fixture uses one value for both IDs, so the 'listed' test passes with the wrong predicate. It does not test real behaviour. Give the fixture a name-based id so the test fails until the predicate matches on ssm_target. (website/src/test/RemoteCrewPanelLaunchForm.test.tsx, medium)

An automated recommendation, not a review: a maintainer still decides. It approves on its own only when the estimated chance of a defect is at most 12% and the change has at most 400 changed lines. Decider v3c (217fe607) · commit c9645920ad · 166 changed lines. If this recommendation is wrong, reply here: replies are how the approver is evaluated.

@JLargent42
JLargent42 force-pushed the fix/34713-failed-launch-status branch from 98805b6 to 42a11d4 Compare November 11, 2026 03:05
@JLargent42

Copy link
Copy Markdown
Contributor Author

AI review disposition for 98805b6f28e6b8231c9ed4fe14ae1771125122b5, fixed in 42a11d4fdbcf213267c313825807c3186fa7288f:

  • Opus 5.5 and GPT 6.1 (blocking, RemoteCrewPanel.tsx:952, a provisioned instance is not proof it is listed in Your crews): fixed.
    • The "listed in Your crews" line now needs isRegistered(activeJob).
    • An instance that provisioned but never registered (the registry write failed, or a reaped job) falls through to launch_failed_check_crews. That line is reworded so it fits both cases it now covers: "The message below says whether anything from this attempt is still running, and how to remove it." The job's own error carries the recovery steps.
    • The same-settings guard stays in every case.
    • New it.each case "registration, with an instance that never registered". It fails on 98805b6f and passes now. The crew-exists case now uses a registered job.
  • Frontend Tests (1), ruStyle register test: fixed. The ru crew-exists string quoted the formal tab label «Ваши команды». It now says «в списке команд на этой странице». All 12 locale style tests pass locally.

@github-actions github-actions Bot added readiness: checking Automated validation is still running readiness: action required A blocking check or review needs attention and removed readiness: action required A blocking check or review needs attention readiness: checking Automated validation is still running labels Nov 11, 2026
@JLargent42
JLargent42 force-pushed the fix/34713-failed-launch-status branch from 42a11d4 to 7528398 Compare November 11, 2026 04:35
@JLargent42

Copy link
Copy Markdown
Contributor Author

Frontend Tests (3) on 42a11d4f failed the hi register test: हटाएँ is an आप-form imperative. Fixed in 7528398ceda9ded4361eaeff13d7765f44036dc5 by switching to the neutral कैसे हटाना है. All 12 locale style tests now pass locally with I18N_BASE_REF set, so the diff-scoped checks ran. Nothing else changed; all five AI lanes passed on 42a11d4f.

@github-actions github-actions Bot added readiness: checking Automated validation is still running and removed readiness: action required A blocking check or review needs attention labels Nov 11, 2026
@JLargent42
JLargent42 force-pushed the fix/34713-failed-launch-status branch from 7528398 to 1c81cb0 Compare November 11, 2026 05:36
@JLargent42

Copy link
Copy Markdown
Contributor Author

AI review disposition for 7528398ceda9ded4361eaeff13d7765f44036dc5, fixed in 1c81cb0d25e4731d824e4510d54c3f0096669b43:

  • GPT 6.1 (blocking, RemoteCrewPanel.tsx:956, a registered job does not prove the crew still exists): fixed by narrowing. This is the fifth finding in one class: what the failure line claims about leftover resources. Per AGENTS.md, only narrowing or deferral remain at that point, so the line no longer claims anything of its own.
    • The launch_failed_crew_exists string is removed from every catalog, along with its isRegistered branch.
    • Every failure from provisioning on shows one line that points at the job's own message, which says what is left and how to remove it. The job is a snapshot, and a crew it registered can be deleted afterwards, so the message is the authority.
    • The same-settings guard is unchanged, and a pre-provision failure is still the plain Launch retry.
  • No live crew-list check. Joining against the live crew list would be a sixth attempt at stating leftover state from the UI, so this PR doesn't add it.
  • The it.each now covers sign-in-after-register, unregistered instance, and provision failure, all expecting the same line. The screenshot of the sign-in state is re-shot.

@JLargent42

Copy link
Copy Markdown
Contributor Author

UX Review on 1c81cb0d25e4731d824e4510d54c3f0096669b43, evidence corrected:

  • Blocker ("the line points to a message that doesn't exist"): the message exists. My screenshot cropped it out. The shot ended 260 px below the card title, so the card's own error notice ("The crew signed in to Kiro as a different identity than the one you chose. Sign out on the crew to fix it.") was below the frame. All after-shots are now re-captured down to the bottom of the progress card. The code is unchanged.
  • Not claiming "this crew exists" from the job: kept on purpose. GPT 6.1 blocked that claim on 7528398c because a registered job is a snapshot: the crew can be deleted afterwards, and the line would then report a crew that isn't there. The card under the line is the live account of what is left.
  • Hiding "Launch another crew" after a post-provision failure: out of this PR's goal. That button is main's existing same-settings path. Removing it after any failure from provisioning on is a product-shape change, so it is left to a maintainer.
  • I re-ran UX Review once against the corrected evidence.

@github-actions github-actions Bot added readiness: action required A blocking check or review needs attention merge conflict Branch has merge conflicts with its base — author must resolve before merge and removed readiness: checking Automated validation is still running labels Nov 11, 2026
@JLargent42
JLargent42 force-pushed the fix/34713-failed-launch-status branch from 1c81cb0 to c964592 Compare November 12, 2026 21:56
@JLargent42

Copy link
Copy Markdown
Contributor Author

Revision c9645920ada20f19af51761e1a2ceb91e3142fc7, the maintainer-chosen answer to the UX Review block on 1c81cb0d:

  • The line now says what is true at the moment it renders.
    • "The launch did not finish. Its crew is listed in Your crews, where you can fix or delete it." appears only when the job registered and its instance_id is in the live instances list the panel already loads.
    • Anything else gets "The launch did not finish. Check Your crews before launching again."
    • This meets UX's request to say so plainly when a crew is there, and GPT 6.1's two earlier objections: an unregistered instance has no row, and a registered crew can be deleted since.
  • Copy is shortened, per UX's note. Each line is one or two short sentences.
  • Tests: a four-case it.each:
    • registered crew in the live list;
    • registered crew deleted since;
    • unregistered instance;
    • provision failure.
  • Also in this revision: rebased onto main 7a7f070a, with the patch-id unchanged. The retry line now names the Launch button by its translated label, since main translates it.
  • Not from this PR: npm run typecheck on main itself fails at website/src/pages/chat/AssistantMessage.tsx:683, where GitFork is used without an import. This PR doesn't touch that file.

@github-actions github-actions Bot added readiness: checking Automated validation is still running readiness: action required A blocking check or review needs attention and removed readiness: action required A blocking check or review needs attention merge conflict Branch has merge conflicts with its base — author must resolve before merge readiness: checking Automated validation is still running labels Nov 12, 2026
After a launch failed, the status line next to the Launch buttons still
said the crew was launched and offered "Launch another crew" with the
same settings, which would fail the same way.

The line now shows an inline error for any failed launch. When the
launch stopped before provisioning started, nothing was created, so the
same-settings guard is lifted and Launch itself is the retry. From
provisioning on the guard stays, since a relaunch could start a second
billed crew. The line says the crew is listed in Your crews only when
the job registered and its instance is in the live crew list, and
otherwise says to check Your crews before launching again.

Fixes #34713
@JLargent42
JLargent42 force-pushed the fix/34713-failed-launch-status branch from c964592 to db7f6e2 Compare November 12, 2026 23:12
@JLargent42

Copy link
Copy Markdown
Contributor Author

GPT 6.1 finding on c9645920ada20f19af51761e1a2ceb91e3142fc7 (blocking, website/src/pages/settings/RemoteCrewPanel.tsx:959, match the crew's SSM target, not its registry ID): fixed in db7f6e2f4c0aa63a850681dbed735c827666ef83.

  • The "listed in Your crews" predicate now matches i.ssm_target === activeJob.instance_id. The job carries the EC2 id, which a crew row stores as its SSM target. The row's own id comes from the crew name.
  • Sites: the LIVE_CREW fixture in website/src/test/RemoteCrewPanelLaunchForm.test.tsx now has a name-based registry id (kiro-crew-cloud-kc-4d10) that differs from its ssm_target. This is what a real registry holds.
  • Proof: with the old i.id === … predicate, "after a launch that failed during sign-in, with its registered crew in the live list…" fails. On db7f6e2f all 12 tests pass. The related RemoteCrewPanel/Settings/i18n subset (103 files, 1678 tests), eslint and typecheck are all green locally.
  • The GPT lane on db7f6e2f hit the installation's primary rate limit and gave no verdict, so it was re-run once.

@JLargent42

Copy link
Copy Markdown
Contributor Author

GPT 6.1 finding on c9645920ada20f19af51761e1a2ceb91e3142fc7 (RemoteCrewPanel.tsx:959, match the crew SSM target, not its registry id): fixed in db7f6e2f4c0aa63a850681dbed735c827666ef83.

  • The live-list check is now instances.some(i => i.ssm_target === activeJob.instance_id). The job carries the EC2 id, which a crew row holds as ssm_target; the row id is derived from the crew name.
  • The LIVE_CREW fixture now has a registry id (kiro-crew-cloud-kc-4d10) distinct from its EC2 id. The "registered crew in the live list" case fails on c9645920 and passes now.
  • Rebased onto main ee18fa53 (includes the GitFork import fix fix(chat): import the GitFork icon the reply menu uses #36124) in the same push, patch-id unchanged. GPT 6.1 reports no blocking findings on this head.

@github-actions github-actions Bot added readiness: passed Eligible automated validation passed for the current revision and removed readiness: action required A blocking check or review needs attention labels Nov 13, 2026
@KiroCrewCaptain
KiroCrewCaptain enabled auto-merge (squash) November 13, 2026 03:08

@KiroCrewCaptain KiroCrewCaptain left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Tier 1 auto-approve: fix (15 files). Criteria: no conflict, no requested changes, security path denylist clean, design-doc gate clean, security checklist all-NO, no reviewer-lane findings, readiness signal fresh for this head SHA, declared category matches the diff. Category: frontend status fix in the remote crew launch form, so a failed cloud launch no longer reads as a success; adds three translated status lines and tests, with no backend change.

@KiroCrewCaptain
KiroCrewCaptain merged commit 5f92090 into main Nov 13, 2026
96 of 111 checks passed
@KiroCrewCaptain
KiroCrewCaptain deleted the fix/34713-failed-launch-status branch November 13, 2026 03:09
@github-actions github-actions Bot removed the readiness: passed Eligible automated validation passed for the current revision label Nov 13, 2026
@bolichen97

Copy link
Copy Markdown
Contributor

Review status: I read the whole diff and have no code concerns. The change adds a failed-launch message, and it lifts main's same-settings guard only when the launch failed before provisioning. That loosens main and adds no refusal. I am holding the approve for one red check.

The rerun / Re-run Sensitive Change Review helper first failed on an API rate limit. After one rerun it now fails with "No Sensitive Change Review run found for PR 34844" at the current head.

@JLargent42, please push a commit or edit the PR title or body once. That starts a Sensitive Change Review run for this head, and I will approve once the helper is green.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

The status line next to the buttons still says the launch succeeded and offers to repeat it with the same settings, whic

3 participants