Skip to content

Compare

Two results on one task. The same model slug means only the harness varies.

linkstashtask 1.0.0 · ruleset 1.0.0 · scoring 1.0.02 results on this task

A attribution

none

no proof

B attribution

none

no proof

Run metric deltas, B minus A
MetricABB − A
Steps000
Tool calls000
Tool error rate000
Total tokens000
Cost USD000
Duration ms000
Different modelThe harness and the model both differ, so the score gap is not attributable to one.

A

Sample

Fake fixture (linkstash/compliant)

no model

99.3

Spec
100.0
Security
100.0
Maintainability
95.0
12 pass0 fail0 unverified
Open result

B

Sample

Fake fixture (linkstash/regressed)

no model

49.9

Spec
50.0
Security
33.3
Maintainability
90.0
6 pass6 fail0 unverifiedcapped
Open result
Bench deltas, B minus A
BenchABB − A
Overall99.349.9-49.4
Spec100.050.0-50.0
Security100.033.3-66.7
Maintainability95.090.0-5.0
Requirement outcomes for both results
RequirementBenchAB
REQ-API-001Bookmark API route existsspecPASSPASS
REQ-API-002Bookmark API exposes a read handlerspecPASSPASS
REQ-API-003Bookmark API exposes a create handlerspecPASSPASS
REQ-AUTH-001Bookmark routes require an authenticated userspecPASSFAIL
REQ-DB-001Shared data layer existsspecPASSPASS
REQ-DEP-001Zod is declared as a dependencyspecPASSFAIL
REQ-ERR-001Bookmark API returns structured errorsspecPASSFAIL
REQ-SEC-001No shell execution dependencysecurityPASSFAIL
REQ-SEC-002No environment file in the repositorysecurityPASSPASS
REQ-SEC-003No dynamic code evaluationsecurityPASSFAIL
REQ-UI-001Bookmark list component existsspecPASSPASS
REQ-VALID-001Bookmark creation validates inputspecPASSFAIL