feat(tasks): cut task delivery to unified metric results#203
Merged
Conversation
Flip the task_delivery group from the legacy bullet path to the unified metrics system: 14 tasks.* metrics, tasks.closed as the KPI tile, a throughput timeseries drilldown replacing the legacy trend extra. Legacy delivery plumbing stays for the v1 dashboard fallback and the team heatmap feed, which still read the bullet path until those surfaces migrate. Signed-off-by: Aleksandr Barkhatov <pm@aleks.bar>
Histogram views plus Distributions blocks for the three median duration metrics — development time, time to resolution, pickup time. Signed-off-by: Aleksandr Barkhatov <pm@aleks.bar>
Populated histograms lead; distributions with no events for the entity in the period sort to the end so the section never opens on a placeholder. Empty tiles now carry the same header as populated ones and center the message in the chart-height area instead of collapsing into the top-left corner. Signed-off-by: Aleksandr Barkhatov <pm@aleks.bar>
Drop the destructive Alert, warning icon, and "N below peers" title — the section heading and red values already signal severity, and "below" misreads on lower-is-better metrics (cost above the median is worse, not below). Neutral Card container; member rows read "trailing". Metric rows adopt the drilldown treatment: colored value then a muted "· median X", with the middot as its own flex item so the row gap frames it evenly. Signed-off-by: Aleksandr Barkhatov <pm@aleks.bar>
Extract the drilldown's gap formatter (× / signed-% / pp / signed
absolute) into a shared util so the KPI tiles and the needs-attention
list convey the scale of divergence, not just the median. Compact on
the dashboard ("3.5× vs median 3,563", "16× vs median 35 lines"),
verbose in the drilldown — one magnitude computation underneath.
Percentage metrics diverge in points, not a ratio: "-35 pp", never a
"-39%" that misreads as points beside a 90% median (and never a
meaningless "0.6×").
Signed-off-by: Aleksandr Barkhatov <pm@aleks.bar>
The collapsed count of 3 left an orphan slot in the two-column grid. Show two rows before the fold; raise the collapse threshold so an exactly-six list has no dead "Show 0 more" toggle. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Signed-off-by: Aleksandr Barkhatov <pm@aleks.bar>
Every peer surface now paints one scale — the quartile rank from the shared standing derivation. Red means bottom quartile, green means top quartile, being with the pack renders calm. - peerStatusToStatus maps in_pack to neutral, not amber: a mid-pack metric is normal, not a warning. - KPI tiles and the drilldown summary card color by rank, not median side. Median-side split painted ~50% of any cohort red by construction; rank draws the ~25% bottom-quartile line the needs-attention list already uses. - Section grading, group cards, and team rollups carry rank (top/in_pack/bottom/unranked) end to end instead of re-deriving a display status per surface; amber survives only where a section earns it (one bottom below the pattern bar). - Drop medianSide from PeerStanding — no remaining consumers. The unified group-card headline drops the redundant median absolute, keeping "value · gap vs median". Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Signed-off-by: Aleksandr Barkhatov <pm@aleks.bar>
A pp gap computed from raw fractions never added up to the rounded value and median shown beside it (55% · -35.3 pp vs median 90%), reading as a bug on the one gap form users check by addition. Compute the pp spread from the displayed (rounded) operands so the on-screen arithmetic closes, and drop the gap text entirely when rounding collapses it to zero rather than printing "0 pp" beside two equal numbers. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Signed-off-by: Aleksandr Barkhatov <pm@aleks.bar>
The period-over-period pp badge showed one decimal while the peer-gap footer and the relative-change badge round to integer, mixing precision on the same tile. Round pp trend deltas to whole points. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Signed-off-by: Aleksandr Barkhatov <pm@aleks.bar>
Preview rows are the card's fixed identity: keep a configured key with no value on the card as an em dash instead of dropping it, so card shape stays stable across periods. Retune the preview sets — task delivery shows pickup time over reopen rate, git shows code lines over PR cycle time, AI shows accepted lines over acceptance rate. Replace the one-line empty body with a shared shadcn Empty state sized to a populated card, so empty cards keep the grid footprint. Drop the standing badge on empty cards — "no peer data" above "no metrics with data" restated the same absence twice. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Signed-off-by: Aleksandr Barkhatov <pm@aleks.bar>
Histograms self-card like the trend and breakdown charts and drop straight into the chart grid, replacing the single wrapping "Distributions" card and its nested borders. Empty charts render a shared shadcn Empty placeholder sized to the plot area, so an empty chart keeps its neighbours' footprint instead of collapsing. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Signed-off-by: Aleksandr Barkhatov <pm@aleks.bar>
An empty card has no drilldown to open, so render it as a plain div instead of a button — not focusable, no click target, no hover affordance. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Signed-off-by: Aleksandr Barkhatov <pm@aleks.bar>
Put the card header and content on the card's own 12px rhythm instead of the default 24px section gap, and drop the no-op pb-2/pt-0 overrides the older card variant needed. The empty body fills the content region and centers within it, keeping the min-height footprint floor so an empty card still matches a populated card's height. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Signed-off-by: Aleksandr Barkhatov <pm@aleks.bar>
|
Important Review skippedToo many files! This PR contains 52 files, which is 2 over the limit of 50. To get a review, narrow the scope: Upgrade to a paid plan to raise the limit. ⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (52)
You can disable this status message by setting the Use the checkbox below for a quick retry:
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
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
1 participant
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.
Flips the task-delivery dashboard group and its KPI tile to
/v1/metric-results, mirroring the collaboration cutover, plus thepeer-standing and card polish that rode along.
tasks.closedKPI tile now read the unifiedpath; drilldown carries the throughput trend and duration
distributions.
bottom quartile, green = top, in-pack stays calm. Removes the
median-side split that reddened half of every cohort.
displayed operands); section cards regraded base-rate aware; group
previews stabilized with an em-dash for absent keys.
Emptyfor emptycharts and empty group cards (non-interactive).
Closes constructorfabric/insight#1704