Skip to content

live-code-shell measurement: a new-file step - #35759

Merged
teamleaderleo merged 1 commit into
feat-cmux-nextfrom
lane/live-code-measure-new-file
Nov 4, 2026
Merged

teamleaderleo merged 1 commit into
feat-cmux-nextfrom
lane/live-code-measure-new-file

Conversation

@teamleaderleo

@teamleaderleo teamleaderleo commented Nov 4, 2026 •

Copy link
Copy Markdown
Collaborator

Leo (2026-11-04): "Couldn't we architect stuff to make everything more amenable to hot swappage? ... deal with files"

This splits the measurement change out of #35705 so the fleet can run it at that PR's commits: it runs a scripts/measure script only when the script is the same as on feat-cmux-next. The file's bytes match #35705's exactly (blob f47de56c).

scripts/measure/live-code-shell.sh gains a new step 3:

  • It writes Sources/ViewShell/LiveCodeProbeDot.swift, a new view, and edits Badge.swift to use it.
  • It injects naming only Badge.swift, then records new_file_edit (applied, refused, or the same picture) and new_file_edit_to_pixels_s in result.json.
  • The script deletes the probe file when it exits.

The stored-property step becomes step 4. It now counts only refusals that come after it, and it expects the next generation.

With this base's inject.sh, step 3 records refused, because the patch cannot resolve the new type. #35705 is the change that makes it apply. The script changes nothing outside the measurement.

Part of manaflow-ai/cmuxterm-hq#7066.

🤖 Generated with Claude Code
Agent session: agent session


Autofix with [code]smith
Open in [code]smith · Need help on this PR? Tag @codesmith-bot with what you need.


Summary by cubic

Splits the new-file measurement step out of #35705 so the fleet can run it at that PR's commits.

  • Step 3 writes LiveCodeProbeDot.swift, injects naming only Badge.swift, and records new_file_edit and new_file_edit_to_pixels_s in result.json.
  • The stored-property step becomes step 4; it counts refusals after it and expects the next generation.
  • The probe file is deleted when the script exits.

At this base's inject.sh, step 3 records refused; #35705 is the change that makes it apply.

Written for commit d2d8534. Summary will update on new commits.

View guided diff Turn on auto-fix

Step 3 writes a new view file the badge uses and injects naming only Badge.swift; result.json
gains new_file_edit and new_file_edit_to_pixels_s. The stored-property step is now step 4. At this
commit's inject.sh the new step records "refused"; cmux#35705 makes it apply.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Agent-Session: https://cmux.com/deeplink/agent?id=aid-45fddab7397278d38551
@cursor

cursor Bot commented Nov 4, 2026

Copy link
Copy Markdown

Bugbot needs on-demand usage enabled

Bugbot uses usage-based billing for this team and requires on-demand usage to be enabled.

A team admin can enable on-demand usage in the Cursor dashboard.

@github-actions

github-actions Bot commented Nov 4, 2026 •

Copy link
Copy Markdown
Contributor

cmux status (one comment, edited in place; each line is updated by the job that owns it)

Review d2d8534: HQ feed card · gallery before/after runs · checks

App of this head (app.zip, kept 1 day)

gh workflow run reload-build.yml -R manaflow-ai/cmux --ref feat-cmux-next -f tag=pr35759 -f ref=lane/live-code-measure-new-file -f platform=macos -f backend_mode=local

Shipped in:

The feat-cmux-next push run https://github.com/manaflow-ai/cmux/actions/runs/41888017760 at dba5f44 failed: cmux-next checks (god files, concurrency, crash safety, l10n).
Those jobs last passed at 2955a17. This pull request is one of 5 merged since: #35756, #35717, #35759, #35521, #35641.
If the failure is in your change, fix forward on feat-cmux-next. Pull requests run only the tiers their paths reach (docs/ci/cmux-next-tiers.md); label a batch PR full-ci to run them all before merging.

@github-actions github-actions Bot added the risk:normal Risk tier (docs/ci/risk-tiers.md) label Nov 4, 2026
@teamleaderleo

Copy link
Copy Markdown
Collaborator Author

Review (codex, router-review): SHIP at d2d8534

Reviewer: codex gpt-6-astra high, fast tier.

Follow-ups: 3. They land as checklist items in this PR's one follow-up tracking issue in this repo (Follow-ups: #N ..., label follow-up; add to an open one before filing, never one issue per finding), linking this receipt; the lane that opened the PR fixes them forward.

SHIP at d2d8534

Visual verdict: no images; Leo tries the build

  • [P1] [follow-up] [category:reliability] [finding:F1] [outcome:pending] scripts/measure/live-code-shell.sh:26,67 overwrites and unconditionally deletes LiveCodeProbeDot.swift; a dirty checkout or later target commit with that file loses it. Preserve any pre-existing file or reject the run.
  • [P2] [follow-up] [category:performance] [finding:F2] [outcome:pending] scripts/measure/live-code-shell.sh:79 waits the full 30 seconds even after the expected refusal is already logged, adding a fixed delay to every base run. Wait for either redraw or refusal.
  • [P2] [follow-up] [category:tooling] [finding:F3] [outcome:pending] scripts/measure/live-code-shell.sh:76-77 swallows compiler/linker failures and reports new_file_edit: "not applied" after the timeout, producing a successful-looking result indistinguishable from a valid refusal. Record injection failure separately or fail the measurement.

Follow-ups:

  • [P1] [finding:F1] Preserve pre-existing probe files.
  • [P2] [finding:F2] Stop waiting when refusal is observed.
  • [P2] [finding:F3] Distinguish injection failures from refusal.

@blacksmith-sh

blacksmith-sh Bot commented Nov 4, 2026

Copy link
Copy Markdown

Found 1 test failure on Blacksmith runners:

Failure

Test View Logs
test_every_listed_hook_is_one_glaeda_released_with_the_check (main.SigningHooksInSt
epWithGlaedaTests.test_every_listed_hook_is_one_glaeda_released_with_the_check)/
test_every_listed_hook_is_one_glaeda_released_with_the_check (main.SigningHooksInSt
epWithGlaedaTests.test_every_listed_hook_is_one_glaeda_released_with_the_check)
View Logs

Fix with [code]smith
Need help on this PR? Tag @codesmith-bot with what you need.

@cmux-bot cmux-bot Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

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

cmux-bot approves d2d8534: an independent review passed this exact head (router-review receipt by teamleaderleo). A push needs a new review; a later BLOCK at this head dismisses this approval.

@teamleaderleo
teamleaderleo merged commit 66d4699 into feat-cmux-next Nov 4, 2026
132 of 149 checks passed
@teamleaderleo
teamleaderleo deleted the lane/live-code-measure-new-file branch November 4, 2026 18:27
teamleaderleo added a commit that referenced this pull request Nov 4, 2026
…G bytes (#35862)

* live-code-shell measurement: judge a redraw by the badge's rendered width, not PNG bytes

The same pixels encode to PNGs of different bytes, so cmp said "redrew" while every generation
drew the build's badge (#35850). Also: refuse to run over an existing probe file, stop waiting on
the new-file step once a refusal is logged, report an inject.sh failure as such, and record the
widths in result.json. Review follow-ups of #35759.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Agent-Session: https://cmux.com/deeplink/agent?id=aid-45fddab7397278d38551

* live-code-shell measurement: refuse a dangling probe link, report step 4's inject failure, name a missing width

Review findings of #35862 at 803b764.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Agent-Session: https://cmux.com/deeplink/agent?id=aid-45fddab7397278d38551

* live-code-shell measurement: restore only this run's Badge.swift backup; check the probe before building

Review finding of #35862 at fd073c8: a reused OUT's older Badge.swift.orig was copied back on exit.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Agent-Session: https://cmux.com/deeplink/agent?id=aid-45fddab7397278d38551

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>

This branch was previously deployed

1 inactive deployment
ci-routing-agent — d2d85347 Deployed Nov 4, 2026 by teamleaderleo via wait for same-key tree publisher #112654
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

risk:normal Risk tier (docs/ci/risk-tiers.md)

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant