@bendyline/gilde 0.1.53 → 0.1.55

package/data/craftbook-templates/ex/executive-level-review/versions/2.0.7/test.json

@@ -0,0 +1,139 @@
+{
+  "schemaVersion": 1,
+  "title": "Executive-Level Review — analytics expansion decision",
+  "objective": "Test whether the Executive-Level Review craftbook challenges a proposal, compares lower-cost alternatives, and produces an evidence-anchored mode verdict with decision conditions.",
+  "tags": [
+    "workflow",
+    "strategy",
+    "executive-review",
+    "decision"
+  ],
+  "prompt": "Use the Executive-Level Review craftbook to review `source/analytics-proposal.md`. Challenge whether the proposed scope is the best way to reach the outcome; compare at least three alternatives including the two-day manual export and waiting. Produce all craftbook artifacts and a decision-ready `tasks/eval/reviews/executive-level-review.md`. Use exactly one allowed mode verdict, score every required dimension from 0–10 with a factual basis, separate unknowns from evidence, and state explicit approve, pause, and stop conditions. Do not invent customer or financial evidence.",
+  "setup": {
+    "projectName": "Analytics expansion review",
+    "about": "A hermetic executive decision review. The seeded proposal is the complete evidence set.",
+    "missionObjectives": "Decide what, if anything, should be approved now while preserving the fastest credible route to learning.",
+    "files": [
+      {
+        "path": "source/analytics-proposal.md",
+        "content": "# Embedded analytics proposal\n\n## Requested decision\nApprove an eight-week build of customer-facing dashboards with scheduled PDF delivery and CSV export.\n\n## Evidence\nTwelve of 86 active customers requested better reporting in support conversations. Three customers agreed to a design pilot, but none signed a paid commitment. Support currently creates weekly exports for five customers.\n\n## Cost and constraints\n- Estimate: two engineers for 8 weeks.\n- New charting service: $4,000 per month at current volume.\n- Security review is complete; accessibility and data-retention reviews are not.\n- The enterprise renewal decision is in six weeks.\n- A manual scheduled-export workflow can be built in two engineer-days and would test delivery cadence, but not dashboard engagement.\n\n## Unknowns\nWillingness to pay, which metrics matter, PDF accessibility, and whether scheduled delivery or interactive dashboards drives retention.\n"
+      }
+    ],
+    "craftbookParams": {
+      "workPath": "tasks/eval"
+    }
+  },
+  "mocks": [],
+  "success": {
+    "summary": "The named craftbook produces a grounded executive verdict, alternatives, scorecard, and terminal task record.",
+    "deliverables": [
+      {
+        "path": "tasks/eval/reviews/executive-level-review.md",
+        "kind": "markdown-report",
+        "artifact": true,
+        "minBytes": 1400,
+        "checks": [
+          {
+            "kind": "contains",
+            "file": "tasks/eval/reviews/executive-level-review.md",
+            "pattern": "^#{1,3}\\s+Executive verdict\\b[\\s\\S]*(SCOPE EXPANSION|SELECTIVE|HOLD|SCOPE REDUCTION)[\\s\\S]*^#{1,3}\\s+Scorecard\\b[\\s\\S]*^#{1,3}\\s+Alternatives\\b[\\s\\S]*^#{1,3}\\s+Recommendation\\b[\\s\\S]*^#{1,3}\\s+Decision conditions\\b",
+            "flags": "im",
+            "label": "verdict, scorecard, alternatives, recommendation, and conditions"
+          },
+          {
+            "kind": "contains",
+            "file": "tasks/eval/reviews/executive-level-review.md",
+            "pattern": "(user value).*?(strategic fit).*?(evidence strength).*?(feasibility).*?(sequencing).*?(reversibility).*?(operating cost).*?(downside)",
+            "flags": "is",
+            "label": "complete executive scorecard"
+          },
+          {
+            "kind": "valueGrounding",
+            "file": "tasks/eval/reviews/executive-level-review.md",
+            "facts": [
+              {
+                "id": "customer-requests",
+                "required": [
+                  "12\\s+(?:of\\s+86\\s+)?(?:active\\s+)?customers"
+                ]
+              },
+              {
+                "id": "pilot-evidence",
+                "required": [
+                  "3\\s+customers?[^.]{0,60}(pilot|agreed)"
+                ]
+              },
+              {
+                "id": "build-duration",
+                "required": [
+                  "8\\s+weeks?"
+                ]
+              },
+              {
+                "id": "operating-cost",
+                "required": [
+                  "\\$4,?000\\s+per\\s+month"
+                ]
+              },
+              {
+                "id": "cheap-alternative",
+                "required": [
+                  "(two|2)[ -]?(engineer[- ])?days?[\\s\\S]{0,120}(manual|scheduled)[ -]export"
+                ]
+              }
+            ]
+          },
+          {
+            "kind": "citationsResolve",
+            "file": "tasks/eval/reviews/executive-level-review.md",
+            "minCitations": 1
+          }
+        ]
+      }
+    ],
+    "taskNotes": {
+      "minBytes": 160,
+      "checks": [
+        {
+          "kind": "contains",
+          "file": "task-notes.md",
+          "pattern": "\\bDONE\\b[\\s\\S]*tasks/eval/reviews/executive-level-review\\.md",
+          "label": "terminal craftbook note names the review"
+        }
+      ],
+      "requireCraftbookTask": true
+    },
+    "taskGraph": {
+      "requireCraftbookTask": true,
+      "requireTerminalStep": true
+    },
+    "unchangedFixtures": [
+      "source/analytics-proposal.md"
+    ]
+  },
+  "rubric": {
+    "artifact": {
+      "path": "tasks/eval/reviews/executive-level-review.md",
+      "kind": "markdown"
+    },
+    "axes": [
+      {
+        "name": "Premise challenge",
+        "description": "The review tests whether action is warranted now and compares materially different ways to learn or deliver value."
+      },
+      {
+        "name": "Evidence discipline",
+        "description": "Scores and conclusions trace to supplied facts, with unsupported assumptions and missing decisions called out."
+      },
+      {
+        "name": "Executive usefulness",
+        "description": "The verdict is decisive, sequenced, reversible where possible, and paired with clear decision-changing conditions."
+      }
+    ]
+  },
+  "qualityFocus": [
+    "One allowed mode verdict is selected",
+    "All scorecard dimensions have evidence-backed scores",
+    "The two-day alternative and missing willingness-to-pay evidence influence the decision"
+  ]
+}