Understand the failure.
Verify the repair.
Small, reproducible software failures. The broken implementation, the fix that didn’t work, and the one that passed—preserved together.
Explore the cases ↓How results are verified ↗WHAT THE ARCHIVE CONTAINS
100840 executable cases. 20168 are open.
Every case records the implementation that fails, the fix that did not work, and the repair that passed its checks—with recorded outputs and source hashes. This release adds 100840 cases across 20168 failure mechanisms and 254 domains.
The open tier gives you the failure and the unsuccessful fix for one case in every mechanism. The remaining 80672 cases, 5 variants per mechanism, are member-only: the verified repair, its recorded checks, and the full fixture suite are held in the member archive. Read the methodology ↗
A RECORD OF WHAT WENT WRONG
Browse the archive / 100840
Circle skirt drafter: cut radius · case 01
The skirt waist is two seam allowances too big.
Circle skirt drafter: cut radius · case 02
The skirt waist is two seam allowances too big.
Circle skirt drafter: cut radius · case 03
The skirt waist is two seam allowances too big.
Circle skirt drafter: cut radius · case 04
The skirt waist is two seam allowances too big.
Circle skirt drafter: cut radius · case 05
The skirt waist is two seam allowances too big.
Circle skirt drafter: hem circumference radius · case 01
Hem circumference ignores the waist hole.
Circle skirt drafter: hem circumference radius · case 02
Hem circumference ignores the waist hole.
Circle skirt drafter: hem circumference radius · case 03
Hem circumference ignores the waist hole.
Circle skirt drafter: hem circumference radius · case 04
Hem circumference ignores the waist hole.
Circle skirt drafter: hem circumference radius · case 05
Hem circumference ignores the waist hole.
Circle skirt drafter: ease application · case 01
The waistband has no wearing ease.
Circle skirt drafter: ease application · case 02
The waistband has no wearing ease.
Circle skirt drafter: ease application · case 03
The waistband has no wearing ease.
Circle skirt drafter: ease application · case 04
The waistband has no wearing ease.
Circle skirt drafter: ease application · case 05
The waistband has no wearing ease.
Circle skirt drafter: fabric span hem · case 01
The fabric width check ignores the hem allowance and the skirt cannot be cut in one piece.
Circle skirt drafter: fabric span hem · case 02
The fabric width check ignores the hem allowance and the skirt cannot be cut in one piece.
Circle skirt drafter: fabric span hem · case 03
The fabric width check ignores the hem allowance and the skirt cannot be cut in one piece.
Circle skirt drafter: fabric span hem · case 04
The fabric width check ignores the hem allowance and the skirt cannot be cut in one piece.
Circle skirt drafter: fabric span hem · case 05
The fabric width check ignores the hem allowance and the skirt cannot be cut in one piece.
Top-down sock fit calculator: negative ease direction · case 01
Socks come out 10% larger than the foot and sag.
Top-down sock fit calculator: negative ease direction · case 02
Socks come out 10% larger than the foot and sag.
Top-down sock fit calculator: negative ease direction · case 03
Socks come out 10% larger than the foot and sag.
Top-down sock fit calculator: negative ease direction · case 04
Socks come out 10% larger than the foot and sag.
Top-down sock fit calculator: negative ease direction · case 05
Socks come out 10% larger than the foot and sag.
Top-down sock fit calculator: stitch multiple rounding · case 01
A 2x2 rib sock casts on 58 stitches.
Top-down sock fit calculator: stitch multiple rounding · case 02
A 2x2 rib sock casts on 58 stitches.
Top-down sock fit calculator: stitch multiple rounding · case 03
A 2x2 rib sock casts on 58 stitches.
Top-down sock fit calculator: stitch multiple rounding · case 04
A 2x2 rib sock casts on 58 stitches.
Top-down sock fit calculator: stitch multiple rounding · case 05
A 2x2 rib sock casts on 58 stitches.
Top-down sock fit calculator: heel flap share · case 01
Heel flaps are a third of the sock.
Top-down sock fit calculator: heel flap share · case 02
Heel flaps are a third of the sock.
Top-down sock fit calculator: heel flap share · case 03
Heel flaps are a third of the sock.
Top-down sock fit calculator: heel flap share · case 04
Heel flaps are a third of the sock.
Top-down sock fit calculator: heel flap share · case 05
Heel flaps are a third of the sock.
Top-down sock fit calculator: gusset pick-up · case 01
A hole forms at the gusset corner.
Top-down sock fit calculator: gusset pick-up · case 02
A hole forms at the gusset corner.
Top-down sock fit calculator: gusset pick-up · case 03
A hole forms at the gusset corner.
Top-down sock fit calculator: gusset pick-up · case 04
A hole forms at the gusset corner.
Top-down sock fit calculator: gusset pick-up · case 05
A hole forms at the gusset corner.
Dataset split: one group appears on both sides · case 01
one group appears on both sides.
Dataset split: one group appears on both sides · case 02
one group appears on both sides.
Dataset split: one group appears on both sides · case 03
one group appears on both sides.
Dataset split: one group appears on both sides · case 04
one group appears on both sides.
Dataset split: one group appears on both sides · case 05
one group appears on both sides.
Dataset split: groups are alternated by input order despite unequal sizes · case 01
groups are alternated by input order despite unequal sizes.
Dataset split: groups are alternated by input order despite unequal sizes · case 02
groups are alternated by input order despite unequal sizes.
Dataset split: groups are alternated by input order despite unequal sizes · case 03
groups are alternated by input order despite unequal sizes.
Dataset split: groups are alternated by input order despite unequal sizes · case 04
groups are alternated by input order despite unequal sizes.
Dataset split: groups are alternated by input order despite unequal sizes · case 05
groups are alternated by input order despite unequal sizes.
Dataset split: class representation is uneven across group partitions · case 01
class representation is uneven across group partitions.
Dataset split: class representation is uneven across group partitions · case 02
class representation is uneven across group partitions.
Dataset split: class representation is uneven across group partitions · case 03
class representation is uneven across group partitions.
Dataset split: class representation is uneven across group partitions · case 04
class representation is uneven across group partitions.
Dataset split: class representation is uneven across group partitions · case 05
class representation is uneven across group partitions.
Dataset split: cutoff timestamp is included in both partitions · case 01
cutoff timestamp is included in both partitions.
Dataset split: cutoff timestamp is included in both partitions · case 02
cutoff timestamp is included in both partitions.
Dataset split: cutoff timestamp is included in both partitions · case 03
cutoff timestamp is included in both partitions.
Dataset split: cutoff timestamp is included in both partitions · case 04
cutoff timestamp is included in both partitions.
Dataset split: cutoff timestamp is included in both partitions · case 05
cutoff timestamp is included in both partitions.
Dataset split: validation begins inside the embargo window · case 01
validation begins inside the embargo window.
Dataset split: validation begins inside the embargo window · case 02
validation begins inside the embargo window.
Dataset split: validation begins inside the embargo window · case 03
validation begins inside the embargo window.
Dataset split: validation begins inside the embargo window · case 04
validation begins inside the embargo window.
Dataset split: validation begins inside the embargo window · case 05
validation begins inside the embargo window.
Dataset split: overlapping windows are retained as training examples · case 01
overlapping windows are retained as training examples.
Dataset split: overlapping windows are retained as training examples · case 02
overlapping windows are retained as training examples.
Dataset split: overlapping windows are retained as training examples · case 03
overlapping windows are retained as training examples.
Dataset split: overlapping windows are retained as training examples · case 04
overlapping windows are retained as training examples.
Dataset split: overlapping windows are retained as training examples · case 05
overlapping windows are retained as training examples.
Dataset split: rolling validation windows reuse future observations as training · case 01
rolling validation windows reuse future observations as training.
Dataset split: rolling validation windows reuse future observations as training · case 02
rolling validation windows reuse future observations as training.
Dataset split: rolling validation windows reuse future observations as training · case 03
rolling validation windows reuse future observations as training.
Dataset split: rolling validation windows reuse future observations as training · case 04
rolling validation windows reuse future observations as training.
Dataset split: rolling validation windows reuse future observations as training · case 05
rolling validation windows reuse future observations as training.
Dataset split: entity holdout selects a record instead of the whole entity · case 01
entity holdout selects a record instead of the whole entity.
Dataset split: entity holdout selects a record instead of the whole entity · case 02
entity holdout selects a record instead of the whole entity.
Dataset split: entity holdout selects a record instead of the whole entity · case 03
entity holdout selects a record instead of the whole entity.
Dataset split: entity holdout selects a record instead of the whole entity · case 04
entity holdout selects a record instead of the whole entity.
Dataset split: entity holdout selects a record instead of the whole entity · case 05
entity holdout selects a record instead of the whole entity.
Dataset split: session side lookup uses the date instead of the session · case 01
session side lookup uses the date instead of the session.
Dataset split: session side lookup uses the date instead of the session · case 02
session side lookup uses the date instead of the session.
Dataset split: session side lookup uses the date instead of the session · case 03
session side lookup uses the date instead of the session.
Dataset split: session side lookup uses the date instead of the session · case 04
session side lookup uses the date instead of the session.
Dataset split: session side lookup uses the date instead of the session · case 05
session side lookup uses the date instead of the session.
Dataset split: a group appears in multiple cross-validation folds · case 01
a group appears in multiple cross-validation folds.
Dataset split: a group appears in multiple cross-validation folds · case 02
a group appears in multiple cross-validation folds.
Dataset split: a group appears in multiple cross-validation folds · case 03
a group appears in multiple cross-validation folds.
Dataset split: a group appears in multiple cross-validation folds · case 04
a group appears in multiple cross-validation folds.
Dataset split: a group appears in multiple cross-validation folds · case 05
a group appears in multiple cross-validation folds.
Dataset split: domain holdout leaks through a second domain label · case 01
domain holdout leaks through a second domain label.
Dataset split: domain holdout leaks through a second domain label · case 02
domain holdout leaks through a second domain label.
Dataset split: domain holdout leaks through a second domain label · case 03
domain holdout leaks through a second domain label.
Dataset split: domain holdout leaks through a second domain label · case 04
domain holdout leaks through a second domain label.
Dataset split: domain holdout leaks through a second domain label · case 05
domain holdout leaks through a second domain label.
Dataset split: records from one source file are scattered · case 01
records from one source file are scattered.
Dataset split: records from one source file are scattered · case 02
records from one source file are scattered.
Dataset split: records from one source file are scattered · case 03
records from one source file are scattered.
Dataset split: records from one source file are scattered · case 04
records from one source file are scattered.
Dataset split: records from one source file are scattered · case 05
records from one source file are scattered.
INSPECTABLE BY DESIGN
Every result has a runnable source.
Runnable implementations with recorded outputs, source hashes, and explicit contracts. Related variants share a failure mechanism and belong together in evaluation splits.
Read the methodology ↗