@bendyline/gilde 0.1.26 → 0.1.28

package/data/craftbook-templates/ev/evidence-brief/versions/1.0.1/test.json

@@ -0,0 +1,110 @@
+{
+  "schemaVersion": 1,
+  "title": "Decision Evidence Brief smoke eval",
+  "objective": "Self-contained smoke eval for the Decision Evidence Brief craftbook using the corpus generic harness.",
+  "tags": [
+    "corpus"
+  ],
+  "prompt": "We need a decision evidence brief for the returns-desk work. The backgound info is in source/brief.md — have a look and write it up in report.md.",
+  "setup": {
+    "projectName": "Decision Evidence Brief Eval",
+    "about": "Self-contained eval project for evidence-brief. Seeded inputs are under workspace/source or workspace/fixtures; final deliverable is workspace/report.md.",
+    "missionObjectives": "Use the Decision Evidence Brief craftbook/template, read the seeded local fixtures, and write report.md without network calls, real credentials, or live services.",
+    "files": [
+      {
+        "path": "source/brief.md",
+        "content": "# Decision Evidence Brief Eval Brief\n\nClient: Boreal Desk, a home-office accessories company.\nAudience: operations leads who need an artifact they can use this week.\n\nFixed source facts for grounding:\n- The returns desk pilot covered 18 SKUs.\n- Median first response improved from 18 hours to 6 hours.\n- Preventable refund leakage fell from 14.2% to 8.9%.\n- The top unresolved complaint is status silence after photo submission.\n- Required next actions are automated status emails, barcode-exception training, and a weekly Finance exception export.\n\nUse these facts when the task asks for prose, analysis, copy, UI content, or test data. Do not use live web services, real credentials, or current outside data.\n\nCraftbook under test: evidence-brief - Decision Evidence Brief.\n"
+      }
+    ],
+    "worker": {
+      "name": "Jules",
+      "role": "Writer"
+    }
+  },
+  "mocks": [],
+  "success": {
+    "summary": "report.md is a source-grounded markdown artifact with required facts and actions.",
+    "deliverables": [
+      {
+        "path": "report.md",
+        "kind": "markdown-report",
+        "minBytes": 1000,
+        "checks": [
+          {
+            "kind": "contains",
+            "file": "report.md",
+            "pattern": "(?:^|\\n)#{1,3}\\s+\\S",
+            "flags": "i"
+          },
+          {
+            "kind": "contains",
+            "file": "report.md",
+            "pattern": "source/brief\\.md|Brief",
+            "flags": "i",
+            "label": "cite or name the seeded source"
+          },
+          {
+            "kind": "contains",
+            "file": "report.md",
+            "pattern": "18\\b[\\s\\S]{0,30}\\bSKUs",
+            "flags": "i",
+            "label": "include the 18-SKU pilot scope"
+          },
+          {
+            "kind": "contains",
+            "file": "report.md",
+            "pattern": "18\\s*hours|6\\s*hours",
+            "flags": "i"
+          },
+          {
+            "kind": "contains",
+            "file": "report.md",
+            "pattern": "14\\.2%|8\\.9%",
+            "flags": "i"
+          },
+          {
+            "kind": "contains",
+            "file": "report.md",
+            "pattern": "status silence|photo submission",
+            "flags": "i"
+          },
+          {
+            "kind": "contains",
+            "file": "report.md",
+            "pattern": "automated status emails|barcode|Finance",
+            "flags": "i"
+          }
+        ]
+      }
+    ]
+  },
+  "rubric": {
+    "artifact": {
+      "path": "report.md",
+      "kind": "markdown"
+    },
+    "axes": [
+      {
+        "name": "grounding",
+        "description": "Seeded facts appear faithfully and the source is cited."
+      },
+      {
+        "name": "structure",
+        "description": "Sections progress logically for the artifact class."
+      },
+      {
+        "name": "completeness",
+        "description": "Required elements (risks, next actions, scope) are all present."
+      },
+      {
+        "name": "tone",
+        "description": "The register fits the audience and artifact class."
+      }
+    ]
+  },
+  "qualityFocus": [
+    "source-grounded prose",
+    "section structure",
+    "next-action clarity"
+  ]
+}

package/data/craftbook-templates/ex/executive-level-review/versions/2.0.3/craftbook.json

@@ -0,0 +1,433 @@
+{
+  "id": "executive-level-review",
+  "name": "Executive-Level Review",
+  "description": "A strategic, executive-level review of a plan or scope: challenge the premise, weigh alternatives, and score each dimension before committing — the 'should we even build this' pass.",
+  "basedOn": {
+    "name": "gstack",
+    "url": "https://github.com/garrytan/gstack"
+  },
+  "plan": "Review the decision, not just the document. Start from the premise, user value, and evidence; compare real alternatives; then test economics, sequencing, reversibility, operational load, and downside. Scores must be anchored to observed facts, with unknowns called out. The final verdict is one of SCOPE EXPANSION, SELECTIVE, HOLD, or SCOPE REDUCTION and includes explicit conditions for changing it.",
+  "entryStepId": "evidence",
+  "triggers": [
+    "think bigger",
+    "expand scope",
+    "strategy review",
+    "rethink this plan"
+  ],
+  "command": "executive-level-review",
+  "steps": [
+    {
+      "id": "evidence",
+      "name": "Build the executive evidence base",
+      "description": "Extract the proposal, intended outcome, proof, assumptions, constraints, and missing decisions.",
+      "prompt": "Read the plan and the workspace material it cites. Restate the decision being requested, target user and outcome, timing, constraints, dependencies, and claimed evidence. Verify cited project facts directly. Separate evidence, inference, and unsupported assumption. Identify the decision owner, irreversible choices, and the cost of waiting. Do not score or recommend yet.\n\nObservable handoff: write the completed result to `reviews/executive-evidence.md` in the project's artifacts drawer with `write_artifact`. Do not merely describe what the file would contain. Re-read it with `read_artifact` before finishing this phase and repair any incomplete sections.",
+      "suggestedRole": "strategy analyst",
+      "advanceWhen": {
+        "file": "reviews/executive-evidence.md",
+        "minBytes": 800,
+        "sniff": "nonempty",
+        "requireChange": true,
+        "artifact": true,
+        "goto": "challenge"
+      },
+      "gate": {
+        "at": "completion",
+        "checks": [
+          {
+            "kind": "minBytes",
+            "file": "reviews/executive-evidence.md",
+            "bytes": 800,
+            "artifact": true
+          },
+          {
+            "kind": "sniff",
+            "file": "reviews/executive-evidence.md",
+            "sniff": "nonempty",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-evidence.md",
+            "pattern": "^##\\s+Decision",
+            "flags": "im",
+            "label": "Decision section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-evidence.md",
+            "pattern": "^##\\s+Evidence",
+            "flags": "im",
+            "label": "Evidence section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-evidence.md",
+            "pattern": "^##\\s+Assumptions",
+            "flags": "im",
+            "label": "Assumptions section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-evidence.md",
+            "pattern": "^##\\s+Constraints",
+            "flags": "im",
+            "label": "Constraints section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-evidence.md",
+            "pattern": "^##\\s+Unknowns",
+            "flags": "im",
+            "label": "Unknowns section",
+            "artifact": true
+          }
+        ],
+        "onReject": "evidence",
+        "maxAttempts": 3
+      },
+      "next": "challenge"
+    },
+    {
+      "id": "challenge",
+      "name": "Challenge the premise and options",
+      "description": "Compare the proposal with materially different ways to reach the outcome.",
+      "prompt": "Use `reviews/executive-evidence.md` to test whether the problem deserves action now and whether the proposal is the best vehicle. Develop at least three alternatives, including doing less or waiting. Compare value, confidence, time to learning, opportunity cost, reversibility, operating burden, security or compliance exposure, and failure mode. Identify which choice each key assumption favors and the cheapest evidence that would flip the ranking. Before working, open `reviews/executive-evidence.md` with `read_artifact`.\n\nObservable handoff: write the completed result to `reviews/executive-options.md` in the project's artifacts drawer with `write_artifact`. Do not merely describe what the file would contain. Re-read it with `read_artifact` before finishing this phase and repair any incomplete sections.",
+      "suggestedRole": "executive strategist",
+      "consumes": [
+        {
+          "file": "reviews/executive-evidence.md",
+          "artifact": true
+        }
+      ],
+      "advanceWhen": {
+        "file": "reviews/executive-options.md",
+        "minBytes": 1000,
+        "sniff": "nonempty",
+        "requireChange": true,
+        "artifact": true,
+        "goto": "verdict"
+      },
+      "gate": {
+        "at": "completion",
+        "checks": [
+          {
+            "kind": "minBytes",
+            "file": "reviews/executive-options.md",
+            "bytes": 1000,
+            "artifact": true
+          },
+          {
+            "kind": "sniff",
+            "file": "reviews/executive-options.md",
+            "sniff": "nonempty",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-options.md",
+            "pattern": "^##\\s+Premise test",
+            "flags": "im",
+            "label": "Premise test section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-options.md",
+            "pattern": "^##\\s+Alternatives",
+            "flags": "im",
+            "label": "Alternatives section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-options.md",
+            "pattern": "^##\\s+Opportunity cost",
+            "flags": "im",
+            "label": "Opportunity cost section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-options.md",
+            "pattern": "^##\\s+Failure modes",
+            "flags": "im",
+            "label": "Failure modes section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-options.md",
+            "pattern": "^##\\s+Decision-changing evidence",
+            "flags": "im",
+            "label": "decision-changing evidence section",
+            "artifact": true
+          }
+        ],
+        "onReject": "challenge",
+        "maxAttempts": 3
+      },
+      "next": "verdict"
+    },
+    {
+      "id": "verdict",
+      "name": "Write the executive review",
+      "description": "Deliver an evidence-anchored scorecard, verdict, and decision conditions.",
+      "prompt": "Write a decision-ready review using the evidence and options files. Score user value, strategic fit, evidence strength, feasibility, sequencing, reversibility, operating cost, and downside from 0 to 10; for each score cite the basis and say what would make it a 10. Choose exactly one mode verdict and explain the two or three decisive reasons. Recommend what to approve now, what to cut or defer, near-term validation, owners, and explicit go, pause, or stop conditions.\n\nObservable handoff: write the completed result to `reviews/executive-level-review.md` in the project's artifacts drawer with `write_artifact`. Do not merely describe what the file would contain. Re-read it with `read_artifact` before finishing this phase and repair any incomplete sections.",
+      "suggestedRole": "executive reviewer",
+      "advanceWhen": {
+        "file": "reviews/executive-level-review.md",
+        "minBytes": 1400,
+        "sniff": "nonempty",
+        "requireChange": true,
+        "artifact": true,
+        "goto": "evaluate"
+      },
+      "gate": {
+        "at": "completion",
+        "checks": [
+          {
+            "kind": "minBytes",
+            "file": "reviews/executive-level-review.md",
+            "bytes": 1400,
+            "artifact": true
+          },
+          {
+            "kind": "sniff",
+            "file": "reviews/executive-level-review.md",
+            "sniff": "nonempty",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-level-review.md",
+            "pattern": "^##\\s+Executive verdict",
+            "flags": "im",
+            "label": "Executive verdict section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-level-review.md",
+            "pattern": "(?:SCOPE EXPANSION|SELECTIVE|HOLD|SCOPE REDUCTION)",
+            "flags": "im",
+            "label": "allowed mode verdict",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-level-review.md",
+            "pattern": "^##\\s+Scorecard",
+            "flags": "im",
+            "label": "Scorecard section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-level-review.md",
+            "pattern": "^##\\s+Alternatives",
+            "flags": "im",
+            "label": "Alternatives section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-level-review.md",
+            "pattern": "^##\\s+Recommendation",
+            "flags": "im",
+            "label": "Recommendation section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-level-review.md",
+            "pattern": "^##\\s+Decision conditions",
+            "flags": "im",
+            "label": "Decision conditions section",
+            "artifact": true
+          }
+        ],
+        "onReject": "verdict",
+        "maxAttempts": 3
+      },
+      "next": "evaluate"
+    },
+    {
+      "id": "evaluate",
+      "name": "Evaluate the deliverable",
+      "description": "Independently grade the observable deliverable and route it to finish, repair, or user escalation.",
+      "prompt": "Review `reviews/executive-level-review.md`, `reviews/executive-evidence.md`, `reviews/executive-options.md` against every criterion below. Inspect the underlying evidence files named by the workflow; do not grade from the author's summary alone.\n\n1. The review tests whether action is warranted, not merely how to execute the supplied plan.\n2. Scores are individually justified by evidence and state what would improve them.\n3. At least three materially different alternatives, including doing less or waiting, are compared.\n4. Opportunity cost, reversibility, operating burden, and credible failure modes influence the recommendation.\n5. The allowed verdict follows from the scorecard and has explicit approval, deferral, and stop conditions.\n6. Unknowns remain visible and no unsupported fact is presented as evidence.\n\nOpen `reviews/executive-level-review.md`, `reviews/executive-evidence.md`, `reviews/executive-options.md` with `read_artifact`. Write the evidence-backed review to `reviews/executive-level-review-evaluation.md` in the artifacts drawer with `write_artifact`. Give each criterion a PASS or FAIL with a concrete path, excerpt, measurement, or observed behavior. End with exactly `Verdict: PASS` or `Verdict: REVISE`. Then use `advance_task_step` for the active task: PASS routes to `finish`; REVISE routes to `repair` for review rounds 1 through 2, and the 3th REVISE routes to `needs-user`. Never route to finish while a criterion is unmet.",
+      "suggestedRole": "strategy reviewer",
+      "consumes": [
+        {
+          "file": "reviews/executive-level-review.md",
+          "artifact": true
+        },
+        {
+          "file": "reviews/executive-evidence.md",
+          "artifact": true
+        },
+        {
+          "file": "reviews/executive-options.md",
+          "artifact": true
+        }
+      ],
+      "gate": {
+        "at": "completion",
+        "checks": [
+          {
+            "kind": "minBytes",
+            "file": "reviews/executive-level-review-evaluation.md",
+            "bytes": 400,
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-level-review-evaluation.md",
+            "pattern": "Verdict:\\s*(?:PASS|REVISE)",
+            "flags": "i",
+            "label": "explicit PASS or REVISE verdict",
+            "artifact": true
+          }
+        ],
+        "onReject": "evaluate",
+        "maxAttempts": 3
+      },
+      "next": "repair"
+    },
+    {
+      "id": "repair",
+      "name": "Repair the deliverable",
+      "description": "Fix only the concrete gaps from the latest independent review.",
+      "prompt": "Read `reviews/executive-level-review-evaluation.md` with `read_artifact` and repair every failed criterion in `reviews/executive-level-review.md`, `reviews/executive-evidence.md`, `reviews/executive-options.md`. Make changes on each file's declared surface (`write_artifact` for artifact inputs, `write_file` for workspace inputs), not in task notes or a reply. Preserve evidence that already passed. Re-run or re-check anything the reviewer found unproven. Ensure `reviews/executive-level-review.md` is genuinely updated this turn so the repair is observable, then hand it back for independent evaluation.",
+      "suggestedRole": "executive reviewer",
+      "consumes": [
+        {
+          "file": "reviews/executive-level-review-evaluation.md",
+          "artifact": true
+        },
+        {
+          "file": "reviews/executive-level-review.md",
+          "artifact": true
+        },
+        {
+          "file": "reviews/executive-evidence.md",
+          "artifact": true
+        },
+        {
+          "file": "reviews/executive-options.md",
+          "artifact": true
+        }
+      ],
+      "advanceWhen": {
+        "file": "reviews/executive-level-review.md",
+        "minBytes": 1400,
+        "sniff": "nonempty",
+        "requireChange": true,
+        "artifact": true,
+        "goto": "evaluate"
+      },
+      "gate": {
+        "at": "completion",
+        "checks": [
+          {
+            "kind": "minBytes",
+            "file": "reviews/executive-level-review.md",
+            "bytes": 1400,
+            "artifact": true
+          },
+          {
+            "kind": "sniff",
+            "file": "reviews/executive-level-review.md",
+            "sniff": "nonempty",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-level-review.md",
+            "pattern": "^##\\s+Executive verdict",
+            "flags": "im",
+            "label": "Executive verdict section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-level-review.md",
+            "pattern": "(?:SCOPE EXPANSION|SELECTIVE|HOLD|SCOPE REDUCTION)",
+            "flags": "im",
+            "label": "allowed mode verdict",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-level-review.md",
+            "pattern": "^##\\s+Scorecard",
+            "flags": "im",
+            "label": "Scorecard section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-level-review.md",
+            "pattern": "^##\\s+Alternatives",
+            "flags": "im",
+            "label": "Alternatives section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-level-review.md",
+            "pattern": "^##\\s+Recommendation",
+            "flags": "im",
+            "label": "Recommendation section",
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "reviews/executive-level-review.md",
+            "pattern": "^##\\s+Decision conditions",
+            "flags": "im",
+            "label": "Decision conditions section",
+            "artifact": true
+          }
+        ],
+        "onReject": "repair",
+        "maxAttempts": 3
+      },
+      "next": "evaluate"
+    },
+    {
+      "id": "finish",
+      "name": "Finish",
+      "description": "All deterministic and reviewer criteria passed.",
+      "prompt": "The independent review passed. Read `reviews/executive-level-review-evaluation.md` with `read_artifact`, then use `write_task_note` to record a concise DONE summary with the final deliverable paths (`reviews/executive-level-review.md`, `reviews/executive-evidence.md`, `reviews/executive-options.md`) and the evidence that each acceptance criterion passed. Report DONE without starting new work.",
+      "suggestedRole": "project lead",
+      "consumes": [
+        {
+          "file": "reviews/executive-level-review-evaluation.md",
+          "artifact": true
+        }
+      ],
+      "terminal": true
+    },
+    {
+      "id": "needs-user",
+      "name": "Escalate unresolved concerns",
+      "description": "The bounded repair loop ended without a defensible pass.",
+      "prompt": "The deliverable did not pass after 3 review rounds. Do not claim success. Read `reviews/executive-level-review-evaluation.md` with `read_artifact`, then use `write_task_note` to record DONE_WITH_CONCERNS: the unmet criteria, what was attempted, the affected paths, and the smallest user decision or missing input needed to continue.",
+      "suggestedRole": "project lead",
+      "consumes": [
+        {
+          "file": "reviews/executive-level-review-evaluation.md",
+          "artifact": true
+        }
+      ],
+      "terminal": true
+    }
+  ],
+  "version": "2.0.3",
+  "releasedAt": "2026-08-14T00:00:00Z"
+}