refactor(core): remove borderline verdict

apps/web/src/content/docs/docs/evaluation/rubrics.mdx

@@ -112,8 +112,7 @@ score = sum(criterion_score / 10 * weight) / sum(total_weights)
 | Verdict | Score |
 |---------|-------|
 | `pass` | ≥ 0.8 |
-| `borderline` | ≥ 0.6 |
-| `fail` | < 0.6 |
+| `fail` | < 0.8 |
 
 ## Authoring Rubrics
 

apps/web/src/content/docs/docs/guides/human-review.mdx

@@ -80,7 +80,7 @@ The `feedback.json` file is a structured annotation of a single eval run. It rec
     {
       "test_id": "test-feature-alpha",
       "verdict": "acceptable",
-      "notes": "Score is borderline (0.72) but behavior is correct — the grader penalized for different phrasing."
+      "notes": "Score is low (0.72) but behavior is correct — the grader penalized for different phrasing."
     },
     {
       "test_id": "test-retrieval-basic",

examples/features/composite/scripts/safety-gate-aggregator.js

@@ -53,8 +53,6 @@ try {
 
     if (finalScore >= 0.8) {
       verdict = 'pass';
-    } else if (finalScore >= 0.6) {
-      verdict = 'borderline';
     } else {
       verdict = 'fail';
     }

examples/features/document-extraction/evals/confusion-metrics.eval.baseline.jsonl

@@ -1,5 +1,5 @@
 {"timestamp":"2026-02-20T21:38:57.573Z","test_id":"metrics-001","dataset":"dataset-confusion-metrics","score":1,"target":"mock_extractor","scores":[{"name":"header_confusion","type":"code-grader","score":1,"weight":1,"verdict":"pass","details":{"metrics":{"invoice_number":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"invoice_date":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"currency":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"supplier.name":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"importer.name":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"gross_total":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1}},"summary":{"total_tp":6,"total_tn":0,"total_fp":0,"total_fn":0,"macro_precision":1,"macro_recall":1,"macro_f1":1}},"assertions":[{"text":"invoice_number: TP (correct non-empty)","passed":true,"evidence":"TP=6 TN=0 FP=0 FN=0, macro-F1=1.000"},{"text":"invoice_date: TP (correct non-empty)","passed":true},{"text":"currency: TP (correct non-empty)","passed":true},{"text":"supplier.name: TP (correct non-empty)","passed":true},{"text":"importer.name: TP (correct non-empty)","passed":true},{"text":"gross_total: TP (correct non-empty)","passed":true}]}],"assertions":[{"text":"invoice_number: TP (correct non-empty)","passed":true,"evidence":"header_confusion: TP=6 TN=0 FP=0 FN=0, macro-F1=1.000"},{"text":"invoice_date: TP (correct non-empty)","passed":true},{"text":"currency: TP (correct non-empty)","passed":true},{"text":"supplier.name: TP (correct non-empty)","passed":true},{"text":"importer.name: TP (correct non-empty)","passed":true},{"text":"gross_total: TP (correct non-empty)","passed":true}]}
-{"timestamp":"2026-02-20T21:38:57.582Z","test_id":"metrics-003","dataset":"dataset-confusion-metrics","score":0.6666666666666666,"target":"mock_extractor","scores":[{"name":"header_confusion","type":"code-grader","score":0.6666666666666666,"weight":1,"verdict":"borderline","details":{"metrics":{"invoice_number":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"invoice_date":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"currency":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"supplier.name":{"tp":0,"tn":0,"fp":1,"fn":1,"precision":0,"recall":0},"importer.name":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"gross_total":{"tp":0,"tn":0,"fp":1,"fn":1,"precision":0,"recall":0}},"summary":{"total_tp":4,"total_tn":0,"total_fp":2,"total_fn":2,"macro_precision":0.6666666666666666,"macro_recall":0.6666666666666666,"macro_f1":0.6666666666666666}},"assertions":[{"text":"invoice_number: TP (correct non-empty)","passed":true,"evidence":"TP=4 TN=0 FP=2 FN=2, macro-F1=0.667"},{"text":"invoice_date: TP (correct non-empty)","passed":true},{"text":"currency: TP (correct non-empty)","passed":true},{"text":"importer.name: TP (correct non-empty)","passed":true},{"text":"supplier.name: FP+FN (wrong value)","passed":false},{"text":"gross_total: FP+FN (wrong value)","passed":false}]}],"assertions":[{"text":"invoice_number: TP (correct non-empty)","passed":true,"evidence":"header_confusion: TP=4 TN=0 FP=2 FN=2, macro-F1=0.667"},{"text":"invoice_date: TP (correct non-empty)","passed":true},{"text":"currency: TP (correct non-empty)","passed":true},{"text":"importer.name: TP (correct non-empty)","passed":true},{"text":"supplier.name: FP+FN (wrong value)","passed":false},{"text":"gross_total: FP+FN (wrong value)","passed":false}]}
+{"timestamp":"2026-02-20T21:38:57.582Z","test_id":"metrics-003","dataset":"dataset-confusion-metrics","score":0.6666666666666666,"target":"mock_extractor","scores":[{"name":"header_confusion","type":"code-grader","score":0.6666666666666666,"weight":1,"verdict":"fail","details":{"metrics":{"invoice_number":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"invoice_date":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"currency":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"supplier.name":{"tp":0,"tn":0,"fp":1,"fn":1,"precision":0,"recall":0},"importer.name":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"gross_total":{"tp":0,"tn":0,"fp":1,"fn":1,"precision":0,"recall":0}},"summary":{"total_tp":4,"total_tn":0,"total_fp":2,"total_fn":2,"macro_precision":0.6666666666666666,"macro_recall":0.6666666666666666,"macro_f1":0.6666666666666666}},"assertions":[{"text":"invoice_number: TP (correct non-empty)","passed":true,"evidence":"TP=4 TN=0 FP=2 FN=2, macro-F1=0.667"},{"text":"invoice_date: TP (correct non-empty)","passed":true},{"text":"currency: TP (correct non-empty)","passed":true},{"text":"importer.name: TP (correct non-empty)","passed":true},{"text":"supplier.name: FP+FN (wrong value)","passed":false},{"text":"gross_total: FP+FN (wrong value)","passed":false}]}],"assertions":[{"text":"invoice_number: TP (correct non-empty)","passed":true,"evidence":"header_confusion: TP=4 TN=0 FP=2 FN=2, macro-F1=0.667"},{"text":"invoice_date: TP (correct non-empty)","passed":true},{"text":"currency: TP (correct non-empty)","passed":true},{"text":"importer.name: TP (correct non-empty)","passed":true},{"text":"supplier.name: FP+FN (wrong value)","passed":false},{"text":"gross_total: FP+FN (wrong value)","passed":false}]}
 {"timestamp":"2026-02-20T21:38:57.588Z","test_id":"metrics-002","dataset":"dataset-confusion-metrics","score":0.8333333333333334,"target":"mock_extractor","scores":[{"name":"header_confusion","type":"code-grader","score":0.8333333333333334,"weight":1,"verdict":"pass","details":{"metrics":{"invoice_number":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"invoice_date":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"currency":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"supplier.name":{"tp":0,"tn":0,"fp":1,"fn":1,"precision":0,"recall":0},"importer.name":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"gross_total":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1}},"summary":{"total_tp":5,"total_tn":0,"total_fp":1,"total_fn":1,"macro_precision":0.8333333333333334,"macro_recall":0.8333333333333334,"macro_f1":0.8333333333333334}},"assertions":[{"text":"invoice_number: TP (correct non-empty)","passed":true,"evidence":"TP=5 TN=0 FP=1 FN=1, macro-F1=0.833"},{"text":"invoice_date: TP (correct non-empty)","passed":true},{"text":"currency: TP (correct non-empty)","passed":true},{"text":"importer.name: TP (correct non-empty)","passed":true},{"text":"gross_total: TP (correct non-empty)","passed":true},{"text":"supplier.name: FP+FN (wrong value)","passed":false}]}],"assertions":[{"text":"invoice_number: TP (correct non-empty)","passed":true,"evidence":"header_confusion: TP=5 TN=0 FP=1 FN=1, macro-F1=0.833"},{"text":"invoice_date: TP (correct non-empty)","passed":true},{"text":"currency: TP (correct non-empty)","passed":true},{"text":"importer.name: TP (correct non-empty)","passed":true},{"text":"gross_total: TP (correct non-empty)","passed":true},{"text":"supplier.name: FP+FN (wrong value)","passed":false}]}
-{"timestamp":"2026-02-20T21:38:57.641Z","test_id":"metrics-004","dataset":"dataset-confusion-metrics","score":0.6666666666666666,"target":"mock_extractor","scores":[{"name":"header_confusion","type":"code-grader","score":0.6666666666666666,"weight":1,"verdict":"borderline","details":{"metrics":{"invoice_number":{"tp":0,"tn":0,"fp":0,"fn":1,"recall":0},"invoice_date":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"currency":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"supplier.name":{"tp":0,"tn":0,"fp":1,"fn":1,"precision":0,"recall":0},"importer.name":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"gross_total":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1}},"summary":{"total_tp":4,"total_tn":0,"total_fp":1,"total_fn":2,"macro_precision":0.8,"macro_recall":0.6666666666666666,"macro_f1":0.6666666666666666}},"assertions":[{"text":"invoice_date: TP (correct non-empty)","passed":true,"evidence":"TP=4 TN=0 FP=1 FN=2, macro-F1=0.667"},{"text":"currency: TP (correct non-empty)","passed":true},{"text":"importer.name: TP (correct non-empty)","passed":true},{"text":"gross_total: TP (correct non-empty)","passed":true},{"text":"invoice_number: FN (missing)","passed":false},{"text":"supplier.name: FP+FN (wrong value)","passed":false}]}],"assertions":[{"text":"invoice_date: TP (correct non-empty)","passed":true,"evidence":"header_confusion: TP=4 TN=0 FP=1 FN=2, macro-F1=0.667"},{"text":"currency: TP (correct non-empty)","passed":true},{"text":"importer.name: TP (correct non-empty)","passed":true},{"text":"gross_total: TP (correct non-empty)","passed":true},{"text":"invoice_number: FN (missing)","passed":false},{"text":"supplier.name: FP+FN (wrong value)","passed":false}]}
+{"timestamp":"2026-02-20T21:38:57.641Z","test_id":"metrics-004","dataset":"dataset-confusion-metrics","score":0.6666666666666666,"target":"mock_extractor","scores":[{"name":"header_confusion","type":"code-grader","score":0.6666666666666666,"weight":1,"verdict":"fail","details":{"metrics":{"invoice_number":{"tp":0,"tn":0,"fp":0,"fn":1,"recall":0},"invoice_date":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"currency":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"supplier.name":{"tp":0,"tn":0,"fp":1,"fn":1,"precision":0,"recall":0},"importer.name":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"gross_total":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1}},"summary":{"total_tp":4,"total_tn":0,"total_fp":1,"total_fn":2,"macro_precision":0.8,"macro_recall":0.6666666666666666,"macro_f1":0.6666666666666666}},"assertions":[{"text":"invoice_date: TP (correct non-empty)","passed":true,"evidence":"TP=4 TN=0 FP=1 FN=2, macro-F1=0.667"},{"text":"currency: TP (correct non-empty)","passed":true},{"text":"importer.name: TP (correct non-empty)","passed":true},{"text":"gross_total: TP (correct non-empty)","passed":true},{"text":"invoice_number: FN (missing)","passed":false},{"text":"supplier.name: FP+FN (wrong value)","passed":false}]}],"assertions":[{"text":"invoice_date: TP (correct non-empty)","passed":true,"evidence":"header_confusion: TP=4 TN=0 FP=1 FN=2, macro-F1=0.667"},{"text":"currency: TP (correct non-empty)","passed":true},{"text":"importer.name: TP (correct non-empty)","passed":true},{"text":"gross_total: TP (correct non-empty)","passed":true},{"text":"invoice_number: FN (missing)","passed":false},{"text":"supplier.name: FP+FN (wrong value)","passed":false}]}
 {"timestamp":"2026-02-20T21:38:57.649Z","test_id":"metrics-005","dataset":"dataset-confusion-metrics","score":0.5,"target":"mock_extractor","scores":[{"name":"header_confusion","type":"code-grader","score":0.5,"weight":1,"verdict":"fail","details":{"metrics":{"invoice_number":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"invoice_date":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"currency":{"tp":0,"tn":0,"fp":1,"fn":1,"precision":0,"recall":0},"supplier.name":{"tp":0,"tn":0,"fp":1,"fn":1,"precision":0,"recall":0},"importer.name":{"tp":1,"tn":0,"fp":0,"fn":0,"precision":1,"recall":1,"f1":1},"gross_total":{"tp":0,"tn":0,"fp":0,"fn":1,"recall":0}},"summary":{"total_tp":3,"total_tn":0,"total_fp":2,"total_fn":3,"macro_precision":0.6,"macro_recall":0.5,"macro_f1":0.5}},"assertions":[{"text":"invoice_number: TP (correct non-empty)","passed":true,"evidence":"TP=3 TN=0 FP=2 FN=3, macro-F1=0.500"},{"text":"invoice_date: TP (correct non-empty)","passed":true},{"text":"importer.name: TP (correct non-empty)","passed":true},{"text":"currency: FP+FN (wrong value)","passed":false},{"text":"supplier.name: FP+FN (wrong value)","passed":false},{"text":"gross_total: FN (missing)","passed":false}]}],"assertions":[{"text":"invoice_number: TP (correct non-empty)","passed":true,"evidence":"header_confusion: TP=3 TN=0 FP=2 FN=3, macro-F1=0.500"},{"text":"invoice_date: TP (correct non-empty)","passed":true},{"text":"importer.name: TP (correct non-empty)","passed":true},{"text":"currency: FP+FN (wrong value)","passed":false},{"text":"supplier.name: FP+FN (wrong value)","passed":false},{"text":"gross_total: FN (missing)","passed":false}]}

examples/features/tool-trajectory-simple/evals/dataset.eval.baseline.jsonl

@@ -2,6 +2,6 @@
 {"timestamp":"2026-02-20T21:40:23.520Z","test_id":"exact-auth-flow","dataset":"dataset","score":1,"target":"mock_agent","scores":[{"name":"auth-sequence-exact","type":"tool-trajectory","score":1,"weight":1,"verdict":"pass","assertions":[{"text":"Position 0: checkCredentials","passed":true},{"text":"Position 1: generateToken","passed":true},{"text":"Position 2: auditLog","passed":true}]}],"assertions":[{"text":"Position 0: checkCredentials","passed":true},{"text":"Position 1: generateToken","passed":true},{"text":"Position 2: auditLog","passed":true}]}
 {"timestamp":"2026-02-20T21:40:23.526Z","test_id":"in-order-pass","dataset":"dataset","score":1,"target":"mock_agent","scores":[{"name":"workflow-sequence","type":"tool-trajectory","score":1,"weight":1,"verdict":"pass","assertions":[{"text":"Found fetchData at position 0","passed":true},{"text":"Found validateSchema at position 1","passed":true},{"text":"Found transformData at position 2","passed":true},{"text":"Found saveResults at position 3","passed":true}]}],"assertions":[{"text":"Found fetchData at position 0","passed":true},{"text":"Found validateSchema at position 1","passed":true},{"text":"Found transformData at position 2","passed":true},{"text":"Found saveResults at position 3","passed":true}]}
 {"timestamp":"2026-02-20T21:40:23.569Z","test_id":"metrics-check","dataset":"dataset","score":1,"target":"mock_agent","scores":[{"name":"metrics-tools","type":"tool-trajectory","score":1,"weight":1,"verdict":"pass","assertions":[{"text":"getCpuMetrics: called 1 times (required \u22651)","passed":true},{"text":"getMemoryMetrics: called 1 times (required \u22651)","passed":true}]}],"assertions":[{"text":"getCpuMetrics: called 1 times (required \u22651)","passed":true},{"text":"getMemoryMetrics: called 1 times (required \u22651)","passed":true}]}
-{"timestamp":"2026-02-20T21:40:23.579Z","test_id":"partial-match","dataset":"dataset","score":0.6666666666666666,"target":"mock_agent","scores":[{"name":"tool-check","type":"tool-trajectory","score":0.6666666666666666,"weight":1,"verdict":"borderline","assertions":[{"text":"knowledgeSearch: called 2 times (required \u22651)","passed":true},{"text":"documentRetrieve: called 1 times (required \u22651)","passed":true},{"text":"generateReport: called 0 times (required \u22651)","passed":false}]}],"assertions":[{"text":"knowledgeSearch: called 2 times (required \u22651)","passed":true},{"text":"documentRetrieve: called 1 times (required \u22651)","passed":true},{"text":"generateReport: called 0 times (required \u22651)","passed":false}]}
+{"timestamp":"2026-02-20T21:40:23.579Z","test_id":"partial-match","dataset":"dataset","score":0.6666666666666666,"target":"mock_agent","scores":[{"name":"tool-check","type":"tool-trajectory","score":0.6666666666666666,"weight":1,"verdict":"fail","assertions":[{"text":"knowledgeSearch: called 2 times (required \u22651)","passed":true},{"text":"documentRetrieve: called 1 times (required \u22651)","passed":true},{"text":"generateReport: called 0 times (required \u22651)","passed":false}]}],"assertions":[{"text":"knowledgeSearch: called 2 times (required \u22651)","passed":true},{"text":"documentRetrieve: called 1 times (required \u22651)","passed":true},{"text":"generateReport: called 0 times (required \u22651)","passed":false}]}
 {"timestamp":"2026-02-20T21:40:23.599Z","test_id":"exact-args-match","dataset":"dataset","score":1,"target":"mock_agent","scores":[{"name":"arg-validation","type":"tool-trajectory","score":1,"weight":1,"verdict":"pass","assertions":[{"text":"Found search at position 0","passed":true},{"text":"Found get_weather at position 1","passed":true}]}],"assertions":[{"text":"Found search at position 0","passed":true},{"text":"Found get_weather at position 1","passed":true}]}
 {"timestamp":"2026-02-20T21:40:23.624Z","test_id":"skip-args-validation","dataset":"dataset","score":1,"target":"mock_agent","scores":[{"name":"workflow-sequence-only","type":"tool-trajectory","score":1,"weight":1,"verdict":"pass","assertions":[{"text":"Found load_data at position 0","passed":true},{"text":"Found transform at position 1","passed":true},{"text":"Found save_data at position 2","passed":true}]}],"assertions":[{"text":"Found load_data at position 0","passed":true},{"text":"Found transform at position 1","passed":true},{"text":"Found save_data at position 2","passed":true}]}

examples/showcase/cross-repo-sync/evals/ground-truth/cases-to-tests.diff

@@ -217,7 +217,7 @@ index fcc60ea..e183d5c 100644
 +| `evalId` | `string` | Test ID |
  | `dataset` | `string` | EVAL name |
  | `score` | `number` | Final score (0-1) |
- | `verdict` | `string` | pass / borderline / fail |
+ | `verdict` | `string` | pass / fail |
 @@ -135,7 +135,7 @@ cat results.jsonl | jq -s 'group_by(.verdict) | map({verdict: .[0].verdict, coun
  # Average score
  cat results.jsonl | jq -s 'map(.score) | add / length'
@@ -1045,7 +1045,7 @@ index b5737e7..0791c70 100644
 -| [Case Schema](/specification/evalcase-schema/) | Within EVAL.yaml | Individual test case definition |
 +| [Test Schema](/specification/case-schema/) | Within EVAL.yaml | Individual test definition |
  | [Evaluators](/specification/evaluators/) | Referenced | Assessment components |
- | [Verdicts](/specification/verdicts/) | Results | Pass/borderline/fail determination |
+ | [Verdicts](/specification/verdicts/) | Results | Pass/fail determination |
  | [Organization](/specification/organization/) | Directory | File organization patterns |
 @@ -46,7 +46,7 @@ execution:
        type: llm_judge

examples/showcase/cross-repo-sync/evals/ground-truth/schema-field-rename.diff

@@ -521,7 +521,7 @@ index 10e7a1c..fcc60ea 100644
 +| `evalId` | `string` | Case ID |
  | `dataset` | `string` | EVAL name |
  | `score` | `number` | Final score (0-1) |
- | `verdict` | `string` | pass / borderline / fail |
+ | `verdict` | `string` | pass / fail |
 diff --git a/docs/src/content/docs/integration/targets.mdx b/docs/src/content/docs/integration/targets.mdx
 index 68218db..9699aeb 100644
 --- a/docs/src/content/docs/integration/targets.mdx
@@ -1172,7 +1172,7 @@ index 1ae8b33..b5737e7 100644
 -| [Evalcase Schema](/specification/evalcase-schema/) | Within EVAL.yaml | Individual test case definition |
 +| [Case Schema](/specification/evalcase-schema/) | Within EVAL.yaml | Individual test case definition |
  | [Evaluators](/specification/evaluators/) | Referenced | Assessment components |
- | [Verdicts](/specification/verdicts/) | Results | Pass/borderline/fail determination |
+ | [Verdicts](/specification/verdicts/) | Results | Pass/fail determination |
  | [Organization](/specification/organization/) | Directory | File organization patterns |
 @@ -46,9 +46,9 @@ execution:
        type: llm_judge

examples/showcase/evaluator-conformance/README.md

@@ -21,7 +21,7 @@ The harness runs an evaluator N times against a labeled fixture dataset:
 
 It then computes per-fixture metrics:
 
-- **Flip rate** — fraction of runs where the verdict (pass/borderline/fail) differs from the first run
+- **Flip rate** — fraction of runs where the verdict (pass/fail) differs from the first run
 - **Mean / Variance** — statistical summary of scores across runs
 - **Bound violations** — scores outside the expected range for ambiguous fixtures