Repository navigation
[release] v0.128.11 - #8218
Merged
Merged
[release] v0.128.11#8218
Conversation
…feat/m-parity-evaluations-list-export
…oper mode)
The Configuration tab now saves the name and description with
PATCH /evaluations/queues/{id}. A simple queue is an evaluation queue with
the same id and reads its name and description from it, so no API change
is needed. Only users with edit_evaluation_queues see the fields, and only
in beta developer mode; everyone else keeps the read-only view.
Entire-Checkpoint: 01M6S1Z6BKYX85XV483GTXZT8H
… drawer The scenario drawer of the /m run page now shows, for the app output and for each evaluator, a button that opens that step's trace in the trace drawer and the time, tokens and cost of the step's own call. The header gets the old app's tools: full screen, fold all sections, copy the scenario as JSON, and a Form or JSON view. Sections fold one by one. The step metrics come from the scenario's metric rows that the page already reads, and the traces are the ones the page reads in one request per page, so the drawer sends no new request. The Annotate tab keeps its plain layout and its save path. Entire-Checkpoint: 01M6S2RKQEJFM0ZCC4QW91YJK2
Pick up to 4 runs from the Compare picker (search, status filters, Selected n/4), kept in ?compare=a,b as in /w. The header shows each compared run as a colored chip. Overview shows one summary column, radar series and bar per run with the delta to the base run; Scenarios follows each scenario with the compared runs' scenario for the same test case. The compared views load lazily. Shared rules live in @agenta/entities/evaluationRun runCompare.ts. Entire-Checkpoint: 01M6S35TN0BSPNJ8DQ1M8APMXJ
…nd load dialog Entire-Checkpoint: 01M6S40YAT8HWBR7BVCEK78ZPA
…mn menu on the evaluations list Evaluations list in /m: a checkbox per run with select all, a bulk Delete behind edit_evaluation_runs with one confirm and one request, Export CSV of the selected runs (or every run on screen) with the columns on screen, and a Columns menu that hides reference and metric columns per browser. The export header and lines are pure functions in @agenta/entities/evaluationRun (runListExportHeaders, runListExportRow, visibleRunListColumns). The export code loads on the first click. Entire-Checkpoint: 01M6S51RQFBT5JMXK8CWV5RMN2
…arity-evaluations-list-export
…arity-evaluation-focus-drawer
…bles, and budget the app evaluations route The bulk actions now sit next to the count and stick to the scroller's left edge, so a table wider than the screen does not push Delete out of view. Adds the speed budget entry for apps/[app_id]/evaluations at its measured size. Entire-Checkpoint: 01M6S5SSF78G57A5KNABQQ0QZB
…arity-evaluation-compare
…rawer' into feat/m-parity-evaluation-compare
…figuration A compared run's scenario row opens the scenario drawer on that run, so Previous and Next stay in it. The Configuration tab shows one column per compared run with its summary, test set, application and evaluators. Entire-Checkpoint: 01M6S6MG07STWKSH62W1ZQHFFM
…uggestions in the /m playground Entire-Checkpoint: 01M6S77Q00SNXXFCXCHSSV4J54
…he base run Uses the old app's rule (runConfigDiff in @agenta/entities): test set ids and revisions, app, app revision and version, evaluator ids and versions in any order. A differing row is tinted and says Differs. Entire-Checkpoint: 01M6S7DF8GVSXE9TDMBCXKNSHT
…ayground-testset-dialog-url
… test cases and the evaluator draft Entire-Checkpoint: 01M6S91994Q0MSY279H311R0RK
…layground-missing-pieces
…vision in the /m drawer Entire-Checkpoint: 01M6SBTK95919DT44PJ8T4PF06
… the real router Entire-Checkpoint: 01M6SDF7ZKH3WCB68MWQ6PXGX0
Contributor
|
Important Review skippedBot user detected. To trigger a single review, invoke the
 |
feat(frontend): edit a queue's name and description in /m
…s-drawer feat(frontend): show each step's trace and metrics in the /m scenario drawer
feat(frontend): compare evaluation runs on the /m run page
…og-url feat(frontend): drop a file to create a test set from the /m playground load dialog, and keep the evaluator playground state in its address
…t-export feat(frontend): evaluations list selection, bulk delete, CSV export and column menu in /m
feat(frontend): run one step, evaluator run status and prompt input suggestions in the /m playground
Contributor
Author
Website previewPreview URL: https://pr-8218-agenta-website-preview.mahmoud-637.workers.dev Built from |
Contributor
Author
Railway Preview Environment
Dependent tests finished. Post /preview for manual review. Environment absence verified; retained storage requires separate provider accounting. Post |
…and speed budgets - Compared scenarios: number each row of a loaded page on its own, so rows without a test case match the right row of the base run. - Compared scenarios: one compared evaluator no longer fills two base evaluator columns. - Evaluations CSV export: a negative metric such as -0.5 stays a number instead of '-0.5. - Speed budgets for the app endpoints, traces and variants pages.
…uns export A run name such as -001 keeps its formula guard, so a spreadsheet does not read it as -1. Only a metric cell that is a plain negative number is written as a number.
…etrics A categorical value such as -001 stays text and keeps the formula guard.
fix(frontend): v0.128.11 release fixes for compared runs, CSV export and speed budgets
mmabrouk
approved these changes
Nov 7, 2026
mmabrouk
left a comment
mmabrouk
left a comment
Member
There was a problem hiding this comment.
🤖 The AI agent says: Approved on Mahmoud's instruction after the v0.128.11 production deploy (workflow 29 runs 42370656130 for US and 42371555141 for EU) passed the live checks.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
New version v0.128.11 in
Contents
Batch 3, the last batch of the /m developer mode parity work: #8206 (queue name and description edit), #8207 (scenario drawer), #8208 (compare runs), #8210 (playground test set dialog and links; the shared link builder also fixes the /w evaluator playground link), #8212 (evaluations list selection, bulk delete, columns and CSV export; shared selection bar), #8214 (playground "Run one step" menu and evaluator result card). Release fix #8222: speed budgets for three app pages, compare row matching, compare evaluator columns, numeric CSV metrics. All behind beta developer mode.
Final QA
Deploy steps
AGENTA_BETA_DEVELOPER_MODE=truefor live US and live EU in agenta_cloud (web and web-mobile env, like agenta_cloud Enhancement: save variant tabs order #1789 for staging and oss), roll the web pods US first, and check routing with a minted account. Rollback: unset it and roll the web pods.