jev-recipes 0.8.0 → 0.8.2

package/dist/catalog/generated/metadata.js

@@ -20,6 +20,7 @@ export const recipeMetadata = [
                 reason: 'Use action-scope to check whether a single action stays within the requested work.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'action-effects',
@@ -43,6 +44,29 @@ export const recipeMetadata = [
                 reason: 'Use action-scope to check whether the action stays within the work that was requested.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'typesafe-ai/jev',
+                date: '2026-09-27T21:19:58.394Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 20,
+                reviewRate: 0.5,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored new scenario families for the September 2026 update; not independently human-reviewed. Previous inspected holdouts moved to development.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'action-reversibility',
@@ -61,6 +85,23 @@ export const recipeMetadata = [
                 reason: 'Use action-scope to check whether the action stays within the requested work before grading how reversible it is.',
             },
         ],
+        evidence: {
+            kind: 'earlier',
+            label: 'Earlier-version measurement',
+            experimental: true,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: null,
+                split: null,
+                cases: 44,
+                ready: null,
+                reviewRate: null,
+                readyAccuracy: null,
+                failed: null,
+                provenance: [],
+                acceptanceMet: null,
+            },
+        },
     },
     {
         id: 'action-scope',
@@ -78,6 +119,7 @@ export const recipeMetadata = [
                 reason: 'Use instruction-fit to decide whether a particular instruction applies.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'age-appropriateness',
@@ -100,6 +142,7 @@ export const recipeMetadata = [
                 reason: 'Use policy-severity to grade how seriously content breaches a stated policy.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'alert-actionability',
@@ -123,6 +166,7 @@ export const recipeMetadata = [
                 reason: 'Use issue-impact to grade how much a reported problem matters, rather than how well the alert tells you what to do about it.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'answer-consistency',
@@ -140,6 +184,7 @@ export const recipeMetadata = [
                 reason: 'Use verify to assess a claim against evidence rather than another statement.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'answer-coverage',
@@ -158,6 +203,7 @@ export const recipeMetadata = [
                 reason: 'Use answer-relevance to assess how directly the draft addresses the overall request.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'answer-disclosures',
@@ -180,6 +226,7 @@ export const recipeMetadata = [
                 reason: 'Use citation-needed to decide whether a statement requires evidence under your rules.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'answer-grade',
@@ -203,6 +250,7 @@ export const recipeMetadata = [
                 reason: 'Use grounding-level when the criterion is fidelity to a source passage rather than a rubric.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'answer-invalidation',
@@ -220,6 +268,7 @@ export const recipeMetadata = [
                 reason: 'Use cache-match to check reuse for a new question when evidence has not changed.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'answer-relevance',
@@ -237,6 +286,7 @@ export const recipeMetadata = [
                 reason: 'Use answer-coverage to check separate question parts individually.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'answerability',
@@ -255,6 +305,23 @@ export const recipeMetadata = [
                 reason: 'Use answer-coverage after drafting to check whether each question was addressed.',
             },
         ],
+        evidence: {
+            kind: 'earlier',
+            label: 'Earlier-version measurement',
+            experimental: true,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: null,
+                split: null,
+                cases: 50,
+                ready: null,
+                reviewRate: null,
+                readyAccuracy: null,
+                failed: null,
+                provenance: [],
+                acceptanceMet: null,
+            },
+        },
     },
     {
         id: 'appeal-grounds-kind',
@@ -285,6 +352,7 @@ export const recipeMetadata = [
                 reason: 'Use feedback-kind to classify general feedback on a product or service, rather than a formal appeal of a decision.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'appointment-request-kind',
@@ -307,6 +375,7 @@ export const recipeMetadata = [
                 reason: 'Use cancellation-check when you only need to know whether a message cancels something.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'argument-fit',
@@ -324,6 +393,7 @@ export const recipeMetadata = [
                 reason: 'Use tool-fit to check the tool capability before choosing its arguments.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'attempted-step',
@@ -341,6 +411,7 @@ export const recipeMetadata = [
                 reason: 'Use troubleshooting-fit to check whether the procedure fits the symptoms.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'attribution-match',
@@ -380,6 +451,7 @@ export const recipeMetadata = [
                 reason: 'Use reference-resolve to select the referent of an ambiguous expression from supplied candidates.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'audience-fit',
@@ -394,6 +466,7 @@ export const recipeMetadata = [
         related: [
             { id: 'tone-check', reason: 'Use tone-check to evaluate specific writing criteria.' },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'breaking-change-signal',
@@ -416,6 +489,7 @@ export const recipeMetadata = [
                 reason: 'Use instruction-conflict when a change may disagree with an existing documented rule or requirement.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'budget-fit',
@@ -438,6 +512,7 @@ export const recipeMetadata = [
                 reason: 'Use plan-completeness to check that the plan covers its goal; a plan can fit the budget by leaving work out.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'budget-narrative-fit',
@@ -461,6 +536,7 @@ export const recipeMetadata = [
                 reason: 'Use commit-message-fit for the analogous check that a commit message describes the change it accompanies.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'bug-report-completeness',
@@ -483,6 +559,7 @@ export const recipeMetadata = [
                 reason: 'Use requirement-testability to check whether a described expected behavior can be verified.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'buying-intent',
@@ -505,6 +582,29 @@ export const recipeMetadata = [
                 reason: 'Use commitment-strength to grade how firmly a message commits to a stated action.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:26.559Z',
+                split: 'held-out',
+                cases: 25,
+                ready: 24,
+                reviewRate: 0.04,
+                readyAccuracy: 0.958,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 25,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'cache-match',
@@ -522,6 +622,7 @@ export const recipeMetadata = [
                 reason: 'Use query-equivalence to compare questions without assessing a saved answer.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'callback-responsibility',
@@ -546,6 +647,29 @@ export const recipeMetadata = [
             },
             { id: 'followup-timing', reason: 'Use followup-timing for when another contact is wanted.' },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:29.897Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 24,
+                reviewRate: 0.4,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'cancellation-check',
@@ -563,6 +687,7 @@ export const recipeMetadata = [
                 reason: 'Use intent-change to assess broader changes to the current goal.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'care-urgency-wording',
@@ -585,6 +710,7 @@ export const recipeMetadata = [
                 reason: 'Use handoff to decide whether a message should leave the automated flow for a human, such as a nurse line.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'category-fit',
@@ -607,6 +733,7 @@ export const recipeMetadata = [
                 reason: 'Use route to pick the best category from several candidates instead of checking one.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'causal-attribution',
@@ -643,6 +770,7 @@ export const recipeMetadata = [
                 reason: 'Use verify to assess supplied evidence for a causal claim; causal-attribution does not establish causation.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'causal-language-strength',
@@ -673,6 +801,7 @@ export const recipeMetadata = [
                 reason: 'Use uncertainty-expression to grade how hedged a statement is overall, rather than the strength of its causal claim specifically.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'certainty-match',
@@ -687,6 +816,7 @@ export const recipeMetadata = [
         related: [
             { id: 'tone-check', reason: 'Use tone-check to evaluate other explicit writing criteria.' },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'change-meaning',
@@ -704,6 +834,7 @@ export const recipeMetadata = [
                 reason: 'Use answer-invalidation to assess how changed evidence affects a saved claim.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'change-risk',
@@ -726,6 +857,7 @@ export const recipeMetadata = [
                 reason: 'Use action-scope to check whether a change stays within the work that was requested.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'change-window-fit',
@@ -749,6 +881,7 @@ export const recipeMetadata = [
                 reason: 'Use action-scope to check whether an action stays within what was requested, rather than whether its timing is allowed.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'checkers-move',
@@ -772,6 +905,23 @@ export const recipeMetadata = [
                 reason: 'Use take-turn for narrative turn eligibility. Checkers games should supply the known acting player directly without another model call.',
             },
         ],
+        evidence: {
+            kind: 'earlier',
+            label: 'Earlier-version measurement',
+            experimental: true,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: null,
+                split: null,
+                cases: 31,
+                ready: null,
+                reviewRate: null,
+                readyAccuracy: null,
+                failed: null,
+                provenance: [],
+                acceptanceMet: null,
+            },
+        },
     },
     {
         id: 'choose-action',
@@ -812,6 +962,23 @@ export const recipeMetadata = [
                 reason: 'Use step-progress to assess an observed outcome after a move; this recipe recommends a candidate before execution.',
             },
         ],
+        evidence: {
+            kind: 'earlier',
+            label: 'Earlier-version measurement',
+            experimental: true,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: null,
+                split: null,
+                cases: 32,
+                ready: null,
+                reviewRate: null,
+                readyAccuracy: null,
+                failed: null,
+                provenance: [],
+                acceptanceMet: null,
+            },
+        },
     },
     {
         id: 'citation-match',
@@ -828,6 +995,7 @@ export const recipeMetadata = [
         related: [
             { id: 'verify', reason: 'Use verify when each claim already has its own paired evidence.' },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'citation-needed',
@@ -845,6 +1013,7 @@ export const recipeMetadata = [
                 reason: 'Use citation-match to find supporting passages once a citation is needed.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'claim-facets',
@@ -875,6 +1044,7 @@ export const recipeMetadata = [
                 reason: 'Use clarify to decide whether to ask the claimant a follow-up question about the facets this recipe finds missing.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'claim-stance',
@@ -910,6 +1080,23 @@ export const recipeMetadata = [
                 reason: 'Use draft-compare for a preference between two responses under a supplied rubric.',
             },
         ],
+        evidence: {
+            kind: 'earlier',
+            label: 'Earlier-version measurement',
+            experimental: true,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: null,
+                split: null,
+                cases: 53,
+                ready: null,
+                reviewRate: null,
+                readyAccuracy: null,
+                failed: null,
+                provenance: [],
+                acceptanceMet: null,
+            },
+        },
     },
     {
         id: 'clarify',
@@ -928,6 +1115,23 @@ export const recipeMetadata = [
                 reason: 'Use query-specificity to assess how focused the question is, without a requirements list.',
             },
         ],
+        evidence: {
+            kind: 'earlier',
+            label: 'Earlier-version measurement',
+            experimental: true,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: null,
+                split: null,
+                cases: 50,
+                ready: null,
+                reviewRate: null,
+                readyAccuracy: null,
+                failed: null,
+                provenance: [],
+                acceptanceMet: null,
+            },
+        },
     },
     {
         id: 'clause-conflict',
@@ -950,6 +1154,7 @@ export const recipeMetadata = [
                 reason: 'Use evidence-conflict to check whether two factual statements contradict each other rather than whether two requirements can both be met.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'clause-kind',
@@ -972,6 +1177,7 @@ export const recipeMetadata = [
                 reason: 'Use document-role to classify the purpose of a whole document rather than one clause.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'clickbait-level',
@@ -994,6 +1200,7 @@ export const recipeMetadata = [
                 reason: 'Use question-leading to detect questions that push the reader toward a particular answer.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'commit-message-fit',
@@ -1016,6 +1223,7 @@ export const recipeMetadata = [
                 reason: 'Use change-meaning to decide whether an edit to text alters its meaning or is editorial only.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'commitment-strength',
@@ -1038,6 +1246,7 @@ export const recipeMetadata = [
                 reason: 'Use certainty-match to compare expressed certainty with the evidence rather than commitment to act.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'comparable-fit',
@@ -1061,6 +1270,7 @@ export const recipeMetadata = [
                 reason: 'Use passage-compare to pick which of two passages better answers a question, rather than to judge similarity between two property descriptions.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'completion-gate',
@@ -1087,6 +1297,29 @@ export const recipeMetadata = [
                 reason: 'Use result-plausibility to check whether a single tool result is a real answer.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'typesafe-ai/jev',
+                date: '2026-09-27T21:20:01.789Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 30,
+                reviewRate: 0.25,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored new scenario families for the September 2026 update; not independently human-reviewed. Previous inspected holdouts moved to development.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'confirmation-match',
@@ -1104,6 +1337,7 @@ export const recipeMetadata = [
                 reason: 'Use cancellation-check to assess instructions to stop or pause a task.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'consent-request',
@@ -1126,6 +1360,7 @@ export const recipeMetadata = [
                 reason: 'Use promise-check to detect commitments the text makes to the reader rather than permission it asks of them.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'consent-scope-fit',
@@ -1148,6 +1383,7 @@ export const recipeMetadata = [
                 reason: 'Use source-applicability to check whether a policy or source document applies to a situation at all.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'constraint-strength',
@@ -1184,6 +1420,7 @@ export const recipeMetadata = [
                 reason: 'Use instruction-conflict to compare the requirements of two instructions.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'contact-opt-out',
@@ -1211,6 +1448,29 @@ export const recipeMetadata = [
                 reason: 'Use followup-timing for a requested delay or condition for future contact.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:27.401Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 33,
+                reviewRate: 0.175,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'content-facets',
@@ -1233,6 +1493,7 @@ export const recipeMetadata = [
                 reason: 'Use argument-fit to judge whether a specific argument supports a specific claim.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'content-freshness-signal',
@@ -1255,6 +1516,7 @@ export const recipeMetadata = [
                 reason: 'Use freshness-needed to decide whether a question requires current information to answer.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'context-prune',
@@ -1282,6 +1544,29 @@ export const recipeMetadata = [
                 reason: 'Use progress-stall to detect an agent that keeps reprocessing the same context.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: true,
+            measurement: {
+                model: 'typesafe-ai/jev',
+                date: '2026-09-27T21:21:18.930Z',
+                split: 'held-out',
+                cases: 20,
+                ready: 11,
+                reviewRate: 0.45,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored new scenario families for the September 2026 update; not independently human-reviewed. Previous inspected holdouts moved to development.',
+                        cases: 20,
+                    },
+                ],
+                acceptanceMet: false,
+            },
+        },
     },
     {
         id: 'context-role',
@@ -1294,6 +1579,7 @@ export const recipeMetadata = [
         ],
         useWhen: 'You want to identify the role a passage plays in answering a question.',
         related: [{ id: 'rerank', reason: 'Use rerank to select and order passages by relevance.' }],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'continuation-compare',
@@ -1317,6 +1603,7 @@ export const recipeMetadata = [
                 reason: 'Use action-compare to compare two next steps against a goal and constraints, rather than two continuations against a musical context.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'correction-target',
@@ -1335,6 +1622,7 @@ export const recipeMetadata = [
                 reason: 'Use reference-resolve for references that are not corrections.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'corrective-action-fit',
@@ -1358,6 +1646,7 @@ export const recipeMetadata = [
                 reason: 'Use troubleshooting-fit to assess a diagnostic procedure rather than a proposed fix.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'daw-request-kind',
@@ -1381,6 +1670,7 @@ export const recipeMetadata = [
                 reason: 'Use turn-intent to classify the conversational role of a turn, such as a question or a correction, rather than which production operation it asks for.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'deadline-risk',
@@ -1404,6 +1694,7 @@ export const recipeMetadata = [
                 reason: 'Use task-complexity to grade how hard a task is before it starts, rather than how it is tracking once underway.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'defect-report-facets',
@@ -1427,6 +1718,7 @@ export const recipeMetadata = [
                 reason: 'Use clarify to phrase the follow-up question once a report is known to be missing a facet.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'delegation-fit',
@@ -1449,6 +1741,7 @@ export const recipeMetadata = [
                 reason: 'Use route to choose among several named delegates at once when more than one might fit.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'diff-hazards',
@@ -1484,6 +1777,23 @@ export const recipeMetadata = [
                 reason: 'Use breaking-change-signal to check whether a change breaks callers.',
             },
         ],
+        evidence: {
+            kind: 'earlier',
+            label: 'Earlier-version measurement',
+            experimental: true,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: null,
+                split: null,
+                cases: 43,
+                ready: null,
+                reviewRate: null,
+                readyAccuracy: null,
+                failed: null,
+                provenance: [],
+                acceptanceMet: null,
+            },
+        },
     },
     {
         id: 'disclosure-facets',
@@ -1507,6 +1817,7 @@ export const recipeMetadata = [
                 reason: "Use clarify to decide whether a buyer's question about the disclosure is too ambiguous to answer, rather than what the disclosure itself addresses.",
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'dispute-kind',
@@ -1530,6 +1841,7 @@ export const recipeMetadata = [
                 reason: 'Use route when the destinations are caller-defined queues rather than this fixed set of billing dispute types.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'document-role',
@@ -1547,6 +1859,7 @@ export const recipeMetadata = [
                 reason: 'Use context-role to assess how a passage relates to a specific question.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'draft-compare',
@@ -1564,6 +1877,7 @@ export const recipeMetadata = [
                 reason: 'Use tone-check to evaluate each writing criterion for one draft.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'dynamic-change',
@@ -1587,6 +1901,7 @@ export const recipeMetadata = [
                 reason: 'Use tone-check to check a draft against caller-defined writing criteria; this recipe steers loudness, not prose.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'eligibility-facets',
@@ -1610,6 +1925,7 @@ export const recipeMetadata = [
                 reason: "Use clarify to decide whether an applicant's question is too ambiguous to answer, rather than which eligibility topics their statement addresses.",
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'emotion-kind',
@@ -1632,6 +1948,7 @@ export const recipeMetadata = [
                 reason: 'Use uncertainty-expression to detect hedging and doubt rather than emotion.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'entity-match',
@@ -1654,6 +1971,29 @@ export const recipeMetadata = [
                 reason: 'Use task-duplicate to catch a task that repeats one already on the list.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:33.689Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 30,
+                reviewRate: 0.25,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'error-acknowledgment',
@@ -1677,6 +2017,7 @@ export const recipeMetadata = [
                 reason: 'Use uncertainty-expression to grade how the message hedges, rather than whether it admits a prior mistake.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'escalation-wording',
@@ -1699,6 +2040,7 @@ export const recipeMetadata = [
                 reason: 'Use handoff to decide whether a conversation should move to a human at all, rather than whether the writer asked for a higher tier.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'evaluation-mention',
@@ -1734,6 +2076,7 @@ export const recipeMetadata = [
                 reason: 'Use uncertainty-expression to label how certain the response sounds about a specific evaluation claim.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'evidence-conflict',
@@ -1751,6 +2094,7 @@ export const recipeMetadata = [
                 reason: 'Use answer-consistency to compare two statements directly.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'evidence-independence',
@@ -1786,6 +2130,7 @@ export const recipeMetadata = [
                 reason: 'Use evidence-conflict to compare what sources say; conflicting reports can still share an origin.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'evidence-novelty',
@@ -1804,6 +2149,7 @@ export const recipeMetadata = [
                 reason: 'Use passage-duplicate to compare overlap between two passages.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'evidence-strength',
@@ -1826,6 +2172,7 @@ export const recipeMetadata = [
                 reason: 'Use answerability to decide whether evidence can answer a whole question.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'exercise-select',
@@ -1849,6 +2196,7 @@ export const recipeMetadata = [
                 reason: 'Use troubleshooting-fit to check whether one supplied procedure addresses a reported problem, rather than to pick the best of several.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'expense-category',
@@ -1872,6 +2220,7 @@ export const recipeMetadata = [
                 reason: 'Use field-select to choose which field of a record holds a value, rather than which category definition a described purchase satisfies.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'explanation-level',
@@ -1895,6 +2244,7 @@ export const recipeMetadata = [
                 reason: 'Use certainty-match to check whether the confidence expressed fits the reasoning given.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'extraction-fidelity',
@@ -1917,6 +2267,29 @@ export const recipeMetadata = [
                 reason: 'Use summary-coverage when the output is prose that should cover the source, not a set of fields.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:32.456Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 40,
+                reviewRate: 0,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'fact-stability',
@@ -1934,6 +2307,7 @@ export const recipeMetadata = [
                 reason: 'Use freshness-needed to decide whether a question needs current information.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'failure-kind',
@@ -1952,6 +2326,7 @@ export const recipeMetadata = [
                 reason: 'Use result-outcome when first determining what a result reports.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'fair-housing-wording',
@@ -1975,6 +2350,7 @@ export const recipeMetadata = [
                 reason: 'Use policy-severity to grade how serious a confirmed violation is against a supplied policy, rather than to detect the wording in the first place.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'feedback-actionability',
@@ -1998,6 +2374,7 @@ export const recipeMetadata = [
                 reason: 'Use tone-check to judge the wording of the same feedback against writing criteria.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'feedback-kind',
@@ -2015,6 +2392,7 @@ export const recipeMetadata = [
                 reason: 'Use turn-intent for the broader purpose of a conversational turn.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'field-select',
@@ -2033,6 +2411,29 @@ export const recipeMetadata = [
                 reason: 'Use reference-resolve to identify what a conversational reference points to.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:31.169Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 30,
+                reviewRate: 0.25,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'financial-advice-signal',
@@ -2056,6 +2457,7 @@ export const recipeMetadata = [
                 reason: "Use certainty-match to check whether a reply's stated confidence fits its evidence, rather than whether it recommends a financial action.",
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'finger-actions',
@@ -2079,6 +2481,7 @@ export const recipeMetadata = [
                 reason: 'Use next-chord to pick the harmony first, then let finger-actions voice it.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'followup-link',
@@ -2097,6 +2500,29 @@ export const recipeMetadata = [
                 reason: 'Use reference-resolve to identify a referenced item rather than an earlier request.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:24.179Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 30,
+                reviewRate: 0.25,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'followup-timing',
@@ -2124,6 +2550,29 @@ export const recipeMetadata = [
                 reason: 'Use callback-responsibility to identify who should initiate a call.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:28.731Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 37,
+                reviewRate: 0.075,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'forecast-confidence-wording',
@@ -2147,6 +2596,7 @@ export const recipeMetadata = [
                 reason: "Use certainty-match to check whether stated confidence fits the supporting evidence, rather than to grade the wording's certainty on its own.",
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'format-fit',
@@ -2169,6 +2619,7 @@ export const recipeMetadata = [
                 reason: 'Use tone-check when the requirement is about voice or register rather than structure.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'freshness-needed',
@@ -2186,6 +2637,7 @@ export const recipeMetadata = [
                 reason: 'Use fact-stability to assess a particular fact rather than a question.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'frustration-signal',
@@ -2203,6 +2655,7 @@ export const recipeMetadata = [
                 reason: 'Use urgency-signal to detect an explicit request for urgent attention.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'funder-fit',
@@ -2226,6 +2679,7 @@ export const recipeMetadata = [
                 reason: 'Use passage-compare for the same pairwise judgment over text passages and a question.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'game-action',
@@ -2249,6 +2703,7 @@ export const recipeMetadata = [
                 reason: 'Use checkers-move for its American/English checkers board format and built-in checkers instructions.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'game-phase',
@@ -2271,6 +2726,7 @@ export const recipeMetadata = [
                 reason: 'Use step-progress to judge how far a non-game task has advanced.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'goal-drift',
@@ -2293,6 +2749,23 @@ export const recipeMetadata = [
                 reason: 'Use intent-change when the user may have redirected the goal, so a step that looks drifted is actually following a new instruction.',
             },
         ],
+        evidence: {
+            kind: 'earlier',
+            label: 'Earlier-version measurement',
+            experimental: true,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: null,
+                split: null,
+                cases: 44,
+                ready: null,
+                reviewRate: null,
+                readyAccuracy: null,
+                failed: null,
+                provenance: [],
+                acceptanceMet: null,
+            },
+        },
     },
     {
         id: 'grounding-level',
@@ -2315,6 +2788,7 @@ export const recipeMetadata = [
                 reason: 'Use answer-relevance to check that the draft addresses the question, not just the sources.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'guest-request-facets',
@@ -2337,6 +2811,7 @@ export const recipeMetadata = [
                 reason: 'Use clarify to decide whether to ask the guest a follow-up question about the facets this recipe finds missing.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'handoff',
@@ -2352,6 +2827,29 @@ export const recipeMetadata = [
         related: [
             { id: 'route', reason: 'Use route to choose a handler when escalation is not the decision.' },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:21.635Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 28,
+                reviewRate: 0.3,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'handoff-completeness',
@@ -2375,6 +2873,7 @@ export const recipeMetadata = [
                 reason: 'Use clarify to find the specific missing requirements once an item grades below complete.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'headline-fit',
@@ -2397,6 +2896,7 @@ export const recipeMetadata = [
                 reason: 'Use attribution-match to check whether a quoted claim is attributed to the source that actually made it.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'incident-match',
@@ -2410,6 +2910,7 @@ export const recipeMetadata = [
         ],
         useWhen: 'You need to connect a support ticket to a supplied known incident.',
         related: [{ id: 'ticket-match', reason: 'Use ticket-match to compare two tickets directly.' }],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'incident-severity-wording',
@@ -2433,6 +2934,7 @@ export const recipeMetadata = [
                 reason: 'Use policy-severity to grade how serious a policy violation is, rather than how serious an operational incident is.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'injection-signal',
@@ -2455,6 +2957,23 @@ export const recipeMetadata = [
                 reason: 'Use instruction-conflict when two legitimate instructions may disagree.',
             },
         ],
+        evidence: {
+            kind: 'earlier',
+            label: 'Earlier-version measurement',
+            experimental: true,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: null,
+                split: null,
+                cases: 44,
+                ready: null,
+                reviewRate: null,
+                readyAccuracy: null,
+                failed: null,
+                provenance: [],
+                acceptanceMet: null,
+            },
+        },
     },
     {
         id: 'instruction-clarity',
@@ -2477,6 +2996,7 @@ export const recipeMetadata = [
                 reason: 'Use clarify to list the specific questions a delegate would need answered when the instruction grades below clear.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'instruction-conflict',
@@ -2499,6 +3019,7 @@ export const recipeMetadata = [
                 reason: 'Use answer-consistency to compare factual claims rather than required behavior.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'instruction-fit',
@@ -2516,6 +3037,7 @@ export const recipeMetadata = [
                 reason: 'Use action-scope to check the boundaries of a proposed action.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'instruction-priority',
@@ -2539,6 +3061,7 @@ export const recipeMetadata = [
                 reason: 'Use priority-compare to order two tasks by importance, rather than to rank two instructions by the authority the policy grants them.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'instruction-readability',
@@ -2568,6 +3091,7 @@ export const recipeMetadata = [
                 reason: 'Use tone-check to judge the register and tone of the instructions rather than their readability.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'instrument-issue-kind',
@@ -2591,6 +3115,7 @@ export const recipeMetadata = [
                 reason: 'Use issue-impact to grade how badly the problem affects the player, rather than what kind of problem it is.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'instrument-report-facets',
@@ -2614,6 +3139,7 @@ export const recipeMetadata = [
                 reason: 'Use bug-report-completeness for software bug reports, where the expected parts are reproduction steps, expected and actual behavior, and environment.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'intake-question-fit',
@@ -2636,6 +3162,7 @@ export const recipeMetadata = [
                 reason: 'Use instruction-fit to check whether a drafted question follows the form-writing instructions you gave.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'intent-change',
@@ -2653,6 +3180,7 @@ export const recipeMetadata = [
                 reason: 'Use cancellation-check for the narrower question of stopping, pausing, or continuing.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'invoice-facets',
@@ -2676,6 +3204,7 @@ export const recipeMetadata = [
                 reason: 'Use field-select to pick which field of a record a value belongs to, rather than to check which fields the invoice text contains at all.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'issue-impact',
@@ -2693,6 +3222,7 @@ export const recipeMetadata = [
                 reason: 'Use urgency-signal to detect requests for immediate attention.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'issue-recurrence',
@@ -2727,6 +3257,7 @@ export const recipeMetadata = [
                 reason: 'Use repeated-attempt to check whether a proposed troubleshooting step repeats an earlier attempt.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'itinerary-feasibility',
@@ -2749,6 +3280,7 @@ export const recipeMetadata = [
                 reason: 'Use clause-conflict when the question is whether two written requirements can both be satisfied, rather than whether a sequence of timed items can be carried out.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'job-post-facets',
@@ -2772,6 +3304,7 @@ export const recipeMetadata = [
                 reason: 'Use requirement-testability to check whether a stated qualification can be verified.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'length-fit',
@@ -2794,6 +3327,7 @@ export const recipeMetadata = [
                 reason: 'Use answer-relevance to check that the response addresses the question at all.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'listener-request-kind',
@@ -2817,6 +3351,7 @@ export const recipeMetadata = [
                 reason: 'Use route when the destinations are caller-defined handlers rather than this fixed set of performance controls.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'listing-compare',
@@ -2839,6 +3374,7 @@ export const recipeMetadata = [
                 reason: 'Use rerank to score many listings independently against one request.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'listing-fact-consistency',
@@ -2862,6 +3398,7 @@ export const recipeMetadata = [
                 reason: 'Use extraction-fidelity to grade how faithfully structured values were pulled from a source document, the reverse direction of this check.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'log-line-kind',
@@ -2885,6 +3422,7 @@ export const recipeMetadata = [
                 reason: 'Use result-outcome to classify how a reported task or job ended, rather than what kind of log line it is.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'loss-cause-kind',
@@ -2907,6 +3445,7 @@ export const recipeMetadata = [
                 reason: 'Use shipment-issue-kind for what went wrong with a delivery; this recipe classifies the peril behind a property or vehicle loss.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'lyric-mood-fit',
@@ -2930,6 +3469,7 @@ export const recipeMetadata = [
                 reason: 'Use certainty-match for the analogous check that the confidence of a statement matches its evidence.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'maintenance-hazard-wording',
@@ -2953,6 +3493,7 @@ export const recipeMetadata = [
                 reason: 'Use issue-impact to grade how badly a reported problem affects the reporter, rather than whether it is a safety hazard.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'medication-mention',
@@ -2975,6 +3516,7 @@ export const recipeMetadata = [
                 reason: 'Use memory-subject to decide whose record a stated medication fact belongs to before storing it.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'memory-relation',
@@ -2992,6 +3534,7 @@ export const recipeMetadata = [
                 reason: 'Use memory-value to assess usefulness before deciding how to store a fact.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'memory-scope',
@@ -3013,6 +3556,7 @@ export const recipeMetadata = [
                 reason: 'Use preference-kind to distinguish ongoing preferences from temporary instructions.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'memory-subject',
@@ -3047,6 +3591,7 @@ export const recipeMetadata = [
                 reason: 'Use preference-kind to distinguish a lasting preference from a temporary instruction.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'memory-value',
@@ -3061,6 +3606,7 @@ export const recipeMetadata = [
         related: [
             { id: 'memory-scope', reason: 'Use memory-scope to determine where a fact applies.' },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'message-facets',
@@ -3083,6 +3629,7 @@ export const recipeMetadata = [
                 reason: 'Use urgency-signal to detect an explicit request for urgent attention.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'methods-facets',
@@ -3106,6 +3653,7 @@ export const recipeMetadata = [
                 reason: 'Use content-facets to label the structural elements of an article, rather than the methodological details of a study.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'model-route',
@@ -3137,6 +3685,29 @@ export const recipeMetadata = [
                 reason: 'Use task-complexity for a five-level complexity grade without picking a model.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:10.135Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 26,
+                reviewRate: 0.35,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'modulation-moment',
@@ -3160,6 +3731,7 @@ export const recipeMetadata = [
                 reason: 'Use step-complete when supplied evidence must establish a stated completion condition, rather than judging a musical seam from the passage itself.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'mood-match',
@@ -3183,6 +3755,7 @@ export const recipeMetadata = [
                 reason: "Use instruction-fit to check whether an instruction's scope covers a task, rather than whether music delivers a feeling.",
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'mood-request',
@@ -3206,6 +3779,7 @@ export const recipeMetadata = [
                 reason: 'Use turn-intent to find out what a message is doing conversationally before asking which mood it requests.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'motivation-source',
@@ -3243,6 +3817,7 @@ export const recipeMetadata = [
                 reason: 'Use claim-stance to label agreement with a specified claim instead of classifying the reason for acting.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'move-explanation-fit',
@@ -3265,6 +3840,7 @@ export const recipeMetadata = [
                 reason: 'Use causal-attribution to judge whether a stated cause is supported by the described outcome outside a game context.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'narrative-consistency',
@@ -3287,6 +3863,7 @@ export const recipeMetadata = [
                 reason: 'Use answer-consistency to compare two discrete statements rather than to scan a single narrative for internal contradictions.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'next-chord',
@@ -3310,6 +3887,7 @@ export const recipeMetadata = [
                 reason: 'Use choose-action when an explicit goal and rulebook govern the pick instead of harmonic continuity in a key and style.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'next-duration',
@@ -3333,6 +3911,7 @@ export const recipeMetadata = [
                 reason: 'Use take-turn to decide whether a player may act at all under game rules, rather than how long the next musical event should last.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'next-note',
@@ -3356,6 +3935,7 @@ export const recipeMetadata = [
                 reason: 'Use choose-action when an explicit goal and rulebook govern the pick instead of melodic continuity in a key and style.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'notice-facets',
@@ -3379,6 +3959,7 @@ export const recipeMetadata = [
                 reason: 'Use response-needed to decide whether a message requires a reply at all, rather than which elements a notice states.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'novelty-claim-level',
@@ -3402,6 +3983,7 @@ export const recipeMetadata = [
                 reason: 'Use outcome-framing to label how a result is framed as gain or loss, rather than how new it is said to be.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'objection-kind',
@@ -3423,6 +4005,7 @@ export const recipeMetadata = [
                 reason: 'Use constraint-strength to grade how binding a stated requirement or limit is.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'objective-fit',
@@ -3446,6 +4029,7 @@ export const recipeMetadata = [
                 reason: 'Use query-specificity to check whether a question identifies a focused information need.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'offer-terms-facets',
@@ -3469,6 +4053,7 @@ export const recipeMetadata = [
                 reason: 'Use invoice-facets for the same presence-check pattern over invoice text instead of offer terms.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'outcome-framing',
@@ -3504,6 +4089,157 @@ export const recipeMetadata = [
                 reason: 'Use choose-action to select among supplied eligible actions; outcome-framing only labels wording.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
+    },
+    {
+        id: 'paragraph-boundary',
+        title: 'Recover a paragraph boundary',
+        description: 'Do two adjacent extracted text fragments continue one paragraph or belong to separate blocks?',
+        category: 'retrieval',
+        tags: ['document', 'ingestion', 'paragraph', 'line-wrap', 'chunking'],
+        limitations: [
+            'Judges adjacent text only; it does not reconstruct reading order, perform OCR, dehyphenate, or merge text.',
+            'Shared subject matter alone does not establish paragraph continuity. Preserve the boundary on review.',
+            'Experimental until document-separated evaluation meets the published accepted-join precision criteria.',
+        ],
+        useWhen: 'Extraction preserved reading order but lost the difference between a hard line wrap and a paragraph break.',
+        related: [
+            {
+                id: 'passage-standalone',
+                reason: 'Use passage-standalone to judge unresolved references within a passage.',
+            },
+            {
+                id: 'topic-shift',
+                reason: 'Use topic-shift for conversational subject changes rather than text boundaries.',
+            },
+            {
+                id: 'text-block-role',
+                reason: 'Use text-block-role to classify an individual block before considering a join.',
+            },
+        ],
+        evidence: {
+            kind: 'public-dataset',
+            label: 'Current public-dataset measurement',
+            experimental: true,
+            measurement: {
+                model: 'typesafe-ai/jev',
+                date: '2026-09-27T21:21:33.447Z',
+                split: 'held-out',
+                cases: 167,
+                ready: 82,
+                reviewRate: 0.509,
+                readyAccuracy: 0.988,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/async_context.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/buffer.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/cli.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/console.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/events.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/fs.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/http.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/https.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/net.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/path.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/perf_hooks.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/querystring.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/readline.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/stream.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/string_decoder.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 7,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/test.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/tty.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/url.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/v8.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/worker_threads.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/zlib.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 8,
+                    },
+                ],
+                acceptanceMet: false,
+            },
+        },
     },
     {
         id: 'passage-compare',
@@ -3526,6 +4262,7 @@ export const recipeMetadata = [
                 reason: 'Use context-role to label what one passage contributes to a question.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'passage-difficulty',
@@ -3549,6 +4286,7 @@ export const recipeMetadata = [
                 reason: 'Use audience-fit to check whether content suits a described audience, rather than to place a passage on a difficulty scale.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'passage-duplicate',
@@ -3566,6 +4304,7 @@ export const recipeMetadata = [
                 reason: 'Use evidence-novelty to compare a passage with the evidence already collected.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'passage-mood',
@@ -3589,6 +4328,7 @@ export const recipeMetadata = [
                 reason: 'Use outcome-framing to label gain and loss wording about an outcome, not the affect of a musical description.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'passage-standalone',
@@ -3611,6 +4351,7 @@ export const recipeMetadata = [
                 reason: 'Use query-specificity to check the other side of retrieval: whether the query is clear enough to match.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'performance-feedback-facets',
@@ -3634,6 +4375,7 @@ export const recipeMetadata = [
                 reason: 'Use tone-check to judge whether feedback meets a tone requirement such as encouraging or direct.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'persuasion-technique',
@@ -3656,6 +4398,7 @@ export const recipeMetadata = [
                 reason: 'Use question-leading to detect a question that steers toward a particular answer.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'phrase-complete',
@@ -3679,6 +4422,7 @@ export const recipeMetadata = [
                 reason: 'Use resolution-check to detect whether a customer reports an issue as resolved, not whether a melody has come to rest.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'pii-presence',
@@ -3701,6 +4445,23 @@ export const recipeMetadata = [
                 reason: 'Use memory-scope to decide how narrowly a fact about a person should be stored.',
             },
         ],
+        evidence: {
+            kind: 'earlier',
+            label: 'Earlier-version measurement',
+            experimental: true,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: null,
+                split: null,
+                cases: 43,
+                ready: null,
+                reviewRate: null,
+                readyAccuracy: null,
+                failed: null,
+                provenance: [],
+                acceptanceMet: null,
+            },
+        },
     },
     {
         id: 'plan-completeness',
@@ -3723,6 +4484,7 @@ export const recipeMetadata = [
                 reason: 'Use clarify when the task itself is ambiguous, since a plan cannot cover requirements the task never made clear.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'policy-compliance',
@@ -3746,6 +4508,7 @@ export const recipeMetadata = [
                 reason: "Use action-scope to check whether an agent's proposed action stays within its permitted scope, rather than whether a submitted expense follows a spending policy.",
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'policy-severity',
@@ -3768,6 +4531,7 @@ export const recipeMetadata = [
                 reason: 'Use promise-check to catch replies that commit beyond what your rules allow.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'politeness-level',
@@ -3790,6 +4554,7 @@ export const recipeMetadata = [
                 reason: 'Use frustration-signal for a categorical read on expressed frustration.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'postmortem-facets',
@@ -3813,6 +4578,7 @@ export const recipeMetadata = [
                 reason: 'Use summary-coverage to check whether a summary covers a source document, rather than whether a document contains expected sections.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'preference-kind',
@@ -3834,6 +4600,7 @@ export const recipeMetadata = [
                 reason: 'Use memory-scope to identify the supported scope of that preference.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'priority-compare',
@@ -3856,6 +4623,7 @@ export const recipeMetadata = [
                 reason: 'Use task-dependency to establish whether one task must finish before the other can start, which is a hard ordering rather than a priority call.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'privacy-notice-facets',
@@ -3878,6 +4646,7 @@ export const recipeMetadata = [
                 reason: 'Use pii-presence to detect personal data in a text rather than statements about how personal data is handled.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'product-match',
@@ -3900,6 +4669,7 @@ export const recipeMetadata = [
                 reason: 'Use tool-fit for the analogous check of whether a tool satisfies a task description.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'progress-stall',
@@ -3922,6 +4692,29 @@ export const recipeMetadata = [
                 reason: 'Use step-progress to grade how much a single observed result moved the objective forward.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:14.022Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 40,
+                reviewRate: 0,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'promise-check',
@@ -3939,6 +4732,29 @@ export const recipeMetadata = [
                 reason: 'Use action-scope to check a proposed action against the requested work.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:25.450Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 40,
+                reviewRate: 0,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'proposal-facets',
@@ -3962,6 +4778,7 @@ export const recipeMetadata = [
                 reason: 'Use content-facets for the structural elements of an article, such as a thesis and evidence, rather than the parts of a funding request.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'qualification-evidence',
@@ -3985,6 +4802,7 @@ export const recipeMetadata = [
                 reason: 'Use verify to check several candidate claims against paired evidence in one call.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'query-equivalence',
@@ -4002,6 +4820,7 @@ export const recipeMetadata = [
                 reason: 'Use cache-match when deciding whether an existing answer can be reused.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'query-specificity',
@@ -4019,6 +4838,7 @@ export const recipeMetadata = [
                 reason: 'Use clarify to check named requirements for missing or ambiguous information.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'question-assumption',
@@ -4052,6 +4872,7 @@ export const recipeMetadata = [
             },
             { id: 'verify', reason: 'Use verify to assess evidential support for the claim itself.' },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'question-leading',
@@ -4096,6 +4917,7 @@ export const recipeMetadata = [
                 reason: 'Use tone-check to assess draft wording against caller-supplied writing criteria.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'question-relevance',
@@ -4119,6 +4941,7 @@ export const recipeMetadata = [
                 reason: 'Use instruction-fit to check whether a written policy or rubric covers a given task.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'reconciliation-match',
@@ -4142,6 +4965,7 @@ export const recipeMetadata = [
                 reason: 'Use ticket-match to decide whether two support tickets report the same issue, rather than whether a ledger entry matches a statement line.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'reference-resolve',
@@ -4160,6 +4984,7 @@ export const recipeMetadata = [
                 reason: 'Use correction-target when the message corrects a particular field or statement.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'repeated-attempt',
@@ -4177,6 +5002,7 @@ export const recipeMetadata = [
                 reason: 'Use step-progress to assess what an attempted step actually changed.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'repetition-level',
@@ -4200,6 +5026,7 @@ export const recipeMetadata = [
                 reason: 'Use progress-stall to detect an agent circling on a task objective, not a melody circling on one figure.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'reply-template-match',
@@ -4215,6 +5042,7 @@ export const recipeMetadata = [
         related: [
             { id: 'route', reason: 'Use route to select a handler rather than a response template.' },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'report-facets',
@@ -4237,6 +5065,7 @@ export const recipeMetadata = [
                 reason: 'Use step-progress to grade how far the reported work moved the objective, rather than what the report contains.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'requirement-testability',
@@ -4268,6 +5097,7 @@ export const recipeMetadata = [
                 reason: 'Use clarify to find missing information across a supplied list of requirements.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'rerank',
@@ -4286,6 +5116,23 @@ export const recipeMetadata = [
                 reason: 'Use answerability to check whether the selected evidence is enough to answer.',
             },
         ],
+        evidence: {
+            kind: 'earlier',
+            label: 'Earlier-version measurement',
+            experimental: true,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: null,
+                split: null,
+                cases: 57,
+                ready: null,
+                reviewRate: null,
+                readyAccuracy: null,
+                failed: null,
+                provenance: [],
+                acceptanceMet: null,
+            },
+        },
     },
     {
         id: 'resolution-check',
@@ -4307,6 +5154,7 @@ export const recipeMetadata = [
                 reason: 'Use step-complete to assess evidence against a supplied completion condition.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'response-needed',
@@ -4324,6 +5172,29 @@ export const recipeMetadata = [
                 reason: 'Use turn-intent to classify the message purpose in more detail.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:22.956Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 40,
+                reviewRate: 0,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'response-refusal',
@@ -4359,6 +5230,23 @@ export const recipeMetadata = [
                 reason: 'Use result-outcome to interpret an observed task result instead of a response claiming to perform it.',
             },
         ],
+        evidence: {
+            kind: 'earlier',
+            label: 'Earlier-version measurement',
+            experimental: true,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: null,
+                split: null,
+                cases: 44,
+                ready: null,
+                reviewRate: null,
+                readyAccuracy: null,
+                failed: null,
+                provenance: [],
+                acceptanceMet: null,
+            },
+        },
     },
     {
         id: 'result-outcome',
@@ -4376,6 +5264,7 @@ export const recipeMetadata = [
                 reason: 'Use step-progress to compare an observation with the previous state.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'result-plausibility',
@@ -4398,6 +5287,7 @@ export const recipeMetadata = [
                 reason: 'Use result-outcome to classify what a result reports happened, such as success or failure, rather than whether it is real.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'result-usefulness',
@@ -4415,6 +5305,7 @@ export const recipeMetadata = [
                 reason: 'Use result-outcome to classify the reported outcome rather than its usefulness.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'retrieval-needed',
@@ -4432,6 +5323,7 @@ export const recipeMetadata = [
                 reason: 'Use freshness-needed to check whether those facts must be current.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'retry-worthwhile',
@@ -4454,6 +5346,29 @@ export const recipeMetadata = [
                 reason: 'Use repeated-attempt to check whether a proposed next attempt is really a fresh approach rather than the same one again.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:15.276Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 39,
+                reviewRate: 0.025,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'review-comment-kind',
@@ -4475,6 +5390,7 @@ export const recipeMetadata = [
                 reason: 'Use turn-intent to classify the communicative purpose of a conversational message.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'review-facets',
@@ -4497,6 +5413,7 @@ export const recipeMetadata = [
                 reason: 'Use issue-impact to grade how badly a reported defect blocks the customer.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'review-recommendation-kind',
@@ -4527,6 +5444,7 @@ export const recipeMetadata = [
                 reason: 'Use feedback-kind to classify general user feedback, rather than the recommendation in a formal peer review.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'review-response-fit',
@@ -4549,6 +5467,7 @@ export const recipeMetadata = [
                 reason: "Use resolution-check to learn whether a customer says a problem is fixed; this recipe judges only whether the business's reply speaks to the review.",
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'rollback-signal',
@@ -4572,6 +5491,7 @@ export const recipeMetadata = [
                 reason: 'Use change-risk before deploying to grade how likely a change is to cause trouble, rather than after the fact to check whether it did.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'root-cause-depth',
@@ -4595,6 +5515,7 @@ export const recipeMetadata = [
                 reason: 'Use explanation-level to grade how much reasoning an answer shows for a conclusion, rather than how far a failure analysis traces its causes.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'route',
@@ -4613,6 +5534,29 @@ export const recipeMetadata = [
                 reason: 'Use turn-intent to identify what a message is doing before choosing a handler.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:19.034Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 24,
+                reviewRate: 0.4,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'route-many',
@@ -4635,6 +5579,29 @@ export const recipeMetadata = [
                 reason: 'Use ticket-match to link a new message to an existing ticket rather than a queue.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:20.238Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 25,
+                reviewRate: 0.375,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'rule-compliance',
@@ -4657,6 +5624,7 @@ export const recipeMetadata = [
                 reason: 'Use action-scope to check whether an agent action stays within a granted permission rather than a game rule set.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'runbook-fit',
@@ -4679,6 +5647,7 @@ export const recipeMetadata = [
                 reason: 'Use source-applicability to check whether a general document applies to a situation, when neither side is an incident or a runbook.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'safety-incident-kind',
@@ -4702,6 +5671,7 @@ export const recipeMetadata = [
                 reason: 'Use failure-kind when the categories are caller-supplied rather than this fixed set of safety event kinds.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'satisfaction-signal',
@@ -4724,6 +5694,7 @@ export const recipeMetadata = [
                 reason: 'Use frustration-signal for a categorical read on expressed frustration in any message.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'search-intent-kind',
@@ -4746,6 +5717,7 @@ export const recipeMetadata = [
                 reason: 'Use query-specificity to judge how narrow or broad a query is rather than what the searcher wants to do.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'sentiment-shift',
@@ -4768,6 +5740,7 @@ export const recipeMetadata = [
                 reason: 'Use resolution-check to decide whether the customer reported the issue resolved.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'settlement-offer-facets',
@@ -4790,6 +5763,7 @@ export const recipeMetadata = [
                 reason: 'Use clarify to decide whether to ask the sender for the elements this recipe finds missing.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'shipment-issue-kind',
@@ -4812,6 +5786,7 @@ export const recipeMetadata = [
                 reason: 'Use issue-impact to grade how badly the reported problem affects the customer rather than what kind of problem it is.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'slot-fit',
@@ -4834,6 +5809,7 @@ export const recipeMetadata = [
                 reason: 'Use step-complete to check whether a scheduling step has finished rather than whether a slot is allowed.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'sound-match',
@@ -4857,6 +5833,7 @@ export const recipeMetadata = [
                 reason: 'Use audience-fit to judge whether content suits a described audience, rather than whether a patch produces a described sound.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'source-applicability',
@@ -4874,6 +5851,7 @@ export const recipeMetadata = [
                 reason: 'Use evidence-conflict to compare two applicable sources for disagreement.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'spam-signal',
@@ -4896,6 +5874,7 @@ export const recipeMetadata = [
                 reason: 'Use response-needed to decide whether a genuine message calls for a reply.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'stage-evidence',
@@ -4918,6 +5897,7 @@ export const recipeMetadata = [
                 reason: 'Use resolution-check to decide whether a support conversation reached resolution.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'step-complete',
@@ -4935,6 +5915,7 @@ export const recipeMetadata = [
                 reason: 'Use result-outcome to interpret a tool result before checking completion.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'step-progress',
@@ -4952,6 +5933,7 @@ export const recipeMetadata = [
                 reason: 'Use step-complete to check whether the completion condition has been met.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'style-kind',
@@ -4975,6 +5957,7 @@ export const recipeMetadata = [
                 reason: 'Use audience-fit to check whether material suits a described audience rather than which style it belongs to.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'summary-coverage',
@@ -4993,6 +5976,7 @@ export const recipeMetadata = [
                 reason: 'Use answer-coverage when the checklist contains questions to answer.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'symptom-facets',
@@ -5015,6 +5999,7 @@ export const recipeMetadata = [
                 reason: 'Use clarify to decide whether to ask the patient a follow-up question about the facets this recipe finds missing.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'take-turn',
@@ -5052,6 +6037,7 @@ export const recipeMetadata = [
                 reason: 'Use response-needed for conversational follow-through, rather than turn and reaction eligibility under game rules.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'task-complexity',
@@ -5071,6 +6057,7 @@ export const recipeMetadata = [
                 reason: 'Use clarify to find missing requirements in a task that grades complex or open-ended.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'task-dependency',
@@ -5093,6 +6080,7 @@ export const recipeMetadata = [
                 reason: 'Use step-complete to check whether evidence establishes a known prerequisite is already satisfied.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'task-duplicate',
@@ -5119,6 +6107,7 @@ export const recipeMetadata = [
                 reason: 'Use ticket-match to compare reported issues rather than requested work.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'task-overlap',
@@ -5141,6 +6130,7 @@ export const recipeMetadata = [
                 reason: 'Use task-dependency to check whether one task must finish before the other can start, which overlap does not imply.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'tempo-change',
@@ -5164,6 +6154,7 @@ export const recipeMetadata = [
                 reason: 'Use step-progress to judge movement toward a caller-defined objective rather than a tempo adjustment.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'tension-level',
@@ -5187,6 +6178,154 @@ export const recipeMetadata = [
                 reason: 'Use step-progress to judge whether a new observation advances a task objective, rather than how far a harmony is from its home chord.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
+    },
+    {
+        id: 'text-block-role',
+        title: 'Identify a text block’s structure',
+        description: 'Is extracted text a heading, body paragraph, list item, code, table, caption, formula, or other block?',
+        category: 'retrieval',
+        tags: ['document', 'ingestion', 'structure', 'block', 'heading'],
+        limitations: [
+            'Classifies supplied text only; it does not read PDFs, perform OCR, recover missing text, or determine reading order.',
+            'Native tags take precedence. Preserve every block when reviewing uncertain structure.',
+            'Experimental until independent document-level evaluation meets the published acceptance criteria.',
+        ],
+        useWhen: 'You have already extracted a text block but its structural markup is missing.',
+        related: [
+            { id: 'document-role', reason: 'Use document-role for the purpose of an entire document.' },
+            {
+                id: 'context-role',
+                reason: 'Use context-role for the contribution a passage makes to a question.',
+            },
+            {
+                id: 'paragraph-boundary',
+                reason: 'Use paragraph-boundary to decide whether adjacent fragments belong to one paragraph.',
+            },
+        ],
+        evidence: {
+            kind: 'public-dataset',
+            label: 'Current public-dataset measurement',
+            experimental: true,
+            measurement: {
+                model: 'typesafe-ai/jev',
+                date: '2026-09-27T21:21:20.879Z',
+                split: 'held-out',
+                cases: 189,
+                ready: 136,
+                reviewRate: 0.28,
+                readyAccuracy: 0.904,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/async_context.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/buffer.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/cli.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/console.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/events.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/fs.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/http.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/https.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/net.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/path.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/perf_hooks.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/querystring.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/readline.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/stream.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/string_decoder.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/test.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/tty.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/url.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/v8.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/worker_threads.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                    {
+                        method: 'public-dataset',
+                        source: 'https://raw.githubusercontent.com/nodejs/node/v22.20.0/doc/api/zlib.md; source Markdown labels, transformed by build-ingestion-cases.mjs; no independent human review; Node.js MIT license.',
+                        cases: 9,
+                    },
+                ],
+                acceptanceMet: false,
+            },
+        },
     },
     {
         id: 'ticket-match',
@@ -5204,6 +6343,7 @@ export const recipeMetadata = [
                 reason: 'Use incident-match to select a known incident for one ticket.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'tone-check',
@@ -5222,6 +6362,7 @@ export const recipeMetadata = [
                 reason: 'Use draft-compare to choose between two drafts under a rubric.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'tool-call-gate',
@@ -5262,6 +6403,29 @@ export const recipeMetadata = [
                 reason: 'Use injection-signal to screen text for embedded instructions before an agent reads it.',
             },
         ],
+        evidence: {
+            kind: 'synthetic',
+            label: 'Current synthetic measurement',
+            experimental: false,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: '2026-09-27T07:56:16.449Z',
+                split: 'held-out',
+                cases: 40,
+                ready: 31,
+                reviewRate: 0.225,
+                readyAccuracy: 1,
+                failed: 0,
+                provenance: [
+                    {
+                        method: 'author-synthetic',
+                        source: 'AI-authored synthetic boundary cases for jev-recipes; labels have not had independent human review.',
+                        cases: 40,
+                    },
+                ],
+                acceptanceMet: true,
+            },
+        },
     },
     {
         id: 'tool-compare',
@@ -5284,6 +6448,7 @@ export const recipeMetadata = [
                 reason: 'Use action-compare to compare two next steps against a goal, rather than two tools against one task.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'tool-fit',
@@ -5301,6 +6466,7 @@ export const recipeMetadata = [
                 reason: 'Use argument-fit to validate the meaning of a proposed tool argument.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'topic-shift',
@@ -5318,6 +6484,7 @@ export const recipeMetadata = [
                 reason: 'Use intent-change when the important question is whether the goal changed.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'trip-purpose-kind',
@@ -5340,6 +6507,7 @@ export const recipeMetadata = [
                 reason: 'Use buying-intent to judge how close the traveler is to booking; this recipe identifies only the reason for the trip.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'troubleshooting-fit',
@@ -5357,6 +6525,7 @@ export const recipeMetadata = [
                 reason: 'Use attempted-step to check whether that procedure has already been tried.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'turn-intent',
@@ -5374,6 +6543,7 @@ export const recipeMetadata = [
                 reason: 'Use intent-change to assess how the message changes an existing goal.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'uncertainty-expression',
@@ -5410,6 +6580,7 @@ export const recipeMetadata = [
                 reason: 'Use verify to check supplied evidence for a claim; expressed certainty is not evidence of truth.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'urgency-signal',
@@ -5427,6 +6598,7 @@ export const recipeMetadata = [
                 reason: 'Use issue-impact to assess the reported practical consequences.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
     {
         id: 'verify',
@@ -5445,6 +6617,23 @@ export const recipeMetadata = [
                 reason: 'Use citation-match to select which passages support one claim.',
             },
         ],
+        evidence: {
+            kind: 'earlier',
+            label: 'Earlier-version measurement',
+            experimental: true,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: null,
+                split: null,
+                cases: 50,
+                ready: null,
+                reviewRate: null,
+                readyAccuracy: null,
+                failed: null,
+                provenance: [],
+                acceptanceMet: null,
+            },
+        },
     },
     {
         id: 'wake-gate',
@@ -5467,6 +6656,23 @@ export const recipeMetadata = [
                 reason: 'Use progress-stall to detect an agent that keeps waking without making progress.',
             },
         ],
+        evidence: {
+            kind: 'earlier',
+            label: 'Earlier-version measurement',
+            experimental: true,
+            measurement: {
+                model: 'jev-1.13.0',
+                date: null,
+                split: null,
+                cases: 44,
+                ready: null,
+                reviewRate: null,
+                readyAccuracy: null,
+                failed: null,
+                provenance: [],
+                acceptanceMet: null,
+            },
+        },
     },
     {
         id: 'workaround-fit',
@@ -5484,5 +6690,6 @@ export const recipeMetadata = [
                 reason: 'Use troubleshooting-fit to assess a diagnostic procedure rather than a workaround.',
             },
         ],
+        evidence: { kind: 'fixture', label: 'Fixture only', experimental: true, measurement: null },
     },
 ];