@bendyline/gilde 0.1.26 → 0.1.28

package/data/craftbook-templates/cr/crossword-forge/versions/1.0.1/test.json

@@ -0,0 +1,135 @@
+{
+  "schemaVersion": 1,
+  "title": "Forge a Crossword — personal words to a verified printable grid",
+  "objective": "Measure whether the crossword-forge craftbook constructs a mechanically-valid interlocking crossword from the seeded family notes — entries readable out of the grid at their coordinates, every entry clued — rendered as a printable page with the solution out of sight.",
+  "tags": [
+    "html-page"
+  ],
+  "prompt": "can you make a crossword for the grandkids with there names and our dog in it? printable pls",
+  "setup": {
+    "projectName": "Loon Lake Puzzle Almanac",
+    "about": "Puzzle almanac project for the Halloran family. Notes about the solvers live at workspace/notes/family.md; finished puzzles are filed on the workspace/puzzles shelf. The working puzzle is workspace/puzzle.json with its printable page at workspace/puzzle.html.",
+    "missionObjectives": "The grid must actually interlock and verify — every entry readable out of the grid at its coordinates, every entry clued. Personalization comes from the family notes. No live services, no real credentials.",
+    "files": [
+      {
+        "path": "notes/family.md",
+        "content": "# The solvers\n\nGrandkids: Nora (12, reigning flashlight-tag champion), Felix (10, dinosaur phase never ended),\nImani (9, reads above her grade and knows it), Theo (7, will need help with the long ones).\n\nThe dog: Biscuit — golden retriever, eats crayons, beloved.\n\nPlaces: the cabin at Loon Lake, the porch swing, Grandpa's garden.\n\nRunning jokes: Grandpa's burnt pancakes, the canoe that tipped twice in one day,\nflashlight tag after dark, \"who fed Biscuit already?\"\n"
+      },
+      {
+        "path": "checks/verify-puzzle.mjs",
+        "content": "// Eval-only re-check: verifies puzzle.json is a real, solvable crossword.\n// Exits 0 and prints PUZZLE OK when every entry reads out of the grid at its\n// coordinates, every entry has a clue, and the entries interlock.\nimport { readFileSync } from 'node:fs';\n\nfunction fail(message) {\n  console.error('PUZZLE BROKEN: ' + message);\n  process.exit(1);\n}\n\nlet puzzle;\ntry {\n  puzzle = JSON.parse(readFileSync('puzzle.json', 'utf8'));\n} catch (err) {\n  fail('puzzle.json missing or not valid JSON (' + err.message + ')');\n}\n\nconst grid = puzzle.grid;\nif (!Array.isArray(grid) || grid.length < 5 || !grid.every((r) => typeof r === 'string')) {\n  fail('grid must be an array of at least 5 row strings');\n}\nconst width = grid[0].length;\nfor (let r = 0; r < grid.length; r++) {\n  if (grid[r].length !== width) fail('row ' + r + ' width differs from row 0');\n  if (!/^[A-Z.]+$/.test(grid[r])) fail('row ' + r + ' has characters outside A-Z and \".\"');\n}\n\nconst entries = puzzle.entries;\nif (!Array.isArray(entries) || entries.length < 4) fail('need at least 4 entries');\n\nconst owners = new Map();\nlet across = 0;\nlet down = 0;\nfor (const e of entries) {\n  const label = typeof e.answer === 'string' && e.answer ? e.answer : 'entry #' + (e.number ?? '?');\n  if (typeof e.answer !== 'string' || !/^[A-Z]{2,}$/.test(e.answer)) fail(label + ': bad answer');\n  if (e.direction !== 'across' && e.direction !== 'down') fail(label + ': bad direction');\n  if (!Number.isInteger(e.row) || !Number.isInteger(e.col)) fail(label + ': bad coordinates');\n  if (typeof e.clue !== 'string' || e.clue.trim().length === 0) fail(label + ': missing clue');\n  if (e.direction === 'across') across++;\n  else down++;\n  for (let i = 0; i < e.answer.length; i++) {\n    const r = e.direction === 'down' ? e.row + i : e.row;\n    const c = e.direction === 'across' ? e.col + i : e.col;\n    if (r < 0 || r >= grid.length || c < 0 || c >= width) fail(label + ' runs off the grid');\n    if (grid[r][c] !== e.answer[i]) {\n      fail(label + ': grid has \"' + grid[r][c] + '\" at ' + r + ',' + c + ' expected \"' + e.answer[i] + '\"');\n    }\n    const key = r + ',' + c;\n    if (!owners.has(key)) owners.set(key, new Set());\n    owners.get(key).add(e.direction + ':' + label);\n  }\n}\nif (across === 0 || down === 0) fail('need both across and down entries');\n\nfor (const e of entries) {\n  let interlocks = false;\n  for (let i = 0; i < e.answer.length && !interlocks; i++) {\n    const r = e.direction === 'down' ? e.row + i : e.row;\n    const c = e.direction === 'across' ? e.col + i : e.col;\n    for (const owner of owners.get(r + ',' + c) ?? []) {\n      if (!owner.startsWith(e.direction + ':')) interlocks = true;\n    }\n  }\n  if (!interlocks) fail(e.answer + ' does not cross any other entry');\n}\n\nconsole.log('PUZZLE OK: ' + entries.length + ' entries (' + across + ' across, ' + down + ' down) verified');\n"
+      }
+    ],
+    "worker": {
+      "name": "Wouter",
+      "role": "Puzzle maker"
+    }
+  },
+  "mocks": [],
+  "success": {
+    "summary": "puzzle.json is a verified interlocking crossword carrying the personal words; puzzle.html is a printable page with clue lists and the solution kept out of sight.",
+    "deliverables": [
+      {
+        "path": "puzzle.json",
+        "kind": "json",
+        "minBytes": 400,
+        "checks": [
+          {
+            "kind": "sniff",
+            "file": "puzzle.json",
+            "sniff": "json-valid"
+          },
+          {
+            "kind": "nodeScriptPasses",
+            "script": "checks/verify-puzzle.mjs",
+            "timeoutMs": 5000,
+            "requiredOutput": [
+              {
+                "pattern": "PUZZLE OK",
+                "label": "grid verifies as solvable"
+              }
+            ]
+          },
+          {
+            "kind": "contains",
+            "file": "puzzle.json",
+            "pattern": "NORA|FELIX|IMANI|THEO",
+            "label": "a grandkid name made the grid"
+          },
+          {
+            "kind": "contains",
+            "file": "puzzle.json",
+            "pattern": "BISCUIT",
+            "label": "the dog made the grid"
+          }
+        ]
+      },
+      {
+        "path": "puzzle.html",
+        "kind": "html-page",
+        "minBytes": 2000,
+        "checks": [
+          {
+            "kind": "cssMinBytes",
+            "bytes": 650,
+            "file": "puzzle.html"
+          },
+          {
+            "kind": "jsParses",
+            "file": "puzzle.html"
+          },
+          {
+            "kind": "contains",
+            "file": "puzzle.html",
+            "pattern": "@page|@media\\s+print",
+            "flags": "i",
+            "label": "print CSS"
+          },
+          {
+            "kind": "contains",
+            "file": "puzzle.html",
+            "pattern": "Across",
+            "flags": "i",
+            "label": "across clue list"
+          },
+          {
+            "kind": "contains",
+            "file": "puzzle.html",
+            "pattern": "Down",
+            "flags": "i",
+            "label": "down clue list"
+          }
+        ]
+      }
+    ]
+  },
+  "rubric": {
+    "artifact": {
+      "path": "puzzle.html",
+      "kind": "html"
+    },
+    "axes": [
+      {
+        "name": "solvability",
+        "description": "The grid interlocks and the clue numbering matches the grid — a solver could actually finish it."
+      },
+      {
+        "name": "personalization",
+        "description": "The names, the dog, and the running jokes from the family notes carry the puzzle; fillers stay in the background."
+      },
+      {
+        "name": "clue craft",
+        "description": "Clues are tuned to the solvers — in-jokes for the personal entries, honest difficulty for the fillers, help-sized for the youngest."
+      },
+      {
+        "name": "print readiness",
+        "description": "The page prints with pencil-sized squares and the solution on its own page, never beside the blank grid."
+      }
+    ]
+  },
+  "qualityFocus": [
+    "verified interlocking grid",
+    "grounded personalization",
+    "solution out of sight"
+  ]
+}

package/data/craftbook-templates/cs/csv-transformer/versions/1.2.2/craftbook.json

@@ -0,0 +1,101 @@
+{
+  "id": "csv-transformer",
+  "name": "CSV / Tabular Transformer",
+  "description": "Reshape and convert tabular data from one schema to another: rename and reorder columns, split or merge fields, derive new columns, pivot/unpivot, change format (CSV<->JSON<->TSV), and filter rows. Locks the input and target schema and the field-by-field mapping FIRST (so the transform is a contract, not guesswork), then builds the transform and emits the output, then verifies row counts and a sample mapping. Use this for reshaping a CSV, converting data formats, remapping columns, flattening JSON to a table, or reformatting an export for another tool.\n\nA gallery craftbook generated from an archetype spec. It runs\n`phase → (per-phase gate) → … → evaluate → (loop) → finish`. Each build\nphase that produces a checkable artifact is followed by a **runtime\ngate-checkpoint** — the runtime verifies the artifact and routes with no\nmodel turn, looping back to redo the phase on a miss. The final `evaluate`\nstep holds a static deliverable gate plus a reviewer QA pass. What it adds\nover the generic `build-loop`: a specialist role per phase, a\ndomain-correct ordering, and a concrete per-phase quality bar.\n\nDeliverables marked \"artifact\" land in the project's artifacts drawer (`write_artifact` / `read_artifact`), not the shipped workspace — review output is not product source.\n\nPhases:\n\n1. Scope the schema mapping (planner) — lock the input schema, target schema, and field-by-field mapping → gated on artifact `notes/schema-scope.md` (markdown-notes)\n2. Build the transform (developer) — apply the mapping and emit the target file → gated on `data/output.csv` (data-file)\n\nThe gates never advance with an unmet criterion, and loop back to the\nowning phase to fix named gaps.\n",
+  "entryStepId": "schema-scope",
+  "triggers": [
+    "reshape a csv",
+    "convert csv to json",
+    "remap columns",
+    "transform tabular data",
+    "reformat a data export"
+  ],
+  "steps": [
+    {
+      "id": "schema-scope",
+      "name": "Scope the schema mapping",
+      "description": "lock the input schema, target schema, and field-by-field mapping",
+      "prompt": "Define the transform contract before writing any code. Step 1: document the INPUT schema (columns, types, format/delimiter) from a sample of the source. Step 2: document the TARGET schema (columns, types, order, output format). Step 3: write the field-by-field mapping table: each target column = which source column(s) and the transform (rename, cast, split on delimiter, concatenate, derive via formula, constant). Step 4: define row-level rules — filters to apply, dedup, how to handle missing source values, and pivot/unpivot if the shape changes. Step 5: write an acceptance-criteria checklist ('output has exactly the target columns in order', 'output parses in the target format', 'row count matches the filter rule', 'a sampled row maps correctly end-to-end', 'no data type errors'). Call write_task_note and write the mapping table + rules + checklist to notes/schema-scope.md.\n\nThe deliverable `notes/schema-scope.md` lands in the project's artifacts drawer — write it with `write_artifact` and read it back with `read_artifact`; the shipped workspace stays untouched.",
+      "suggestedRole": "planner",
+      "advanceWhen": {
+        "file": "notes/schema-scope.md",
+        "minBytes": 1,
+        "sniff": "nonempty",
+        "artifact": true
+      },
+      "gate": {
+        "at": "completion",
+        "checks": [
+          {
+            "kind": "minBytes",
+            "file": "notes/schema-scope.md",
+            "bytes": 120,
+            "artifact": true
+          },
+          {
+            "kind": "sniff",
+            "file": "notes/schema-scope.md",
+            "sniff": "nonempty",
+            "artifact": true
+          }
+        ],
+        "onReject": "schema-scope",
+        "maxAttempts": 3
+      },
+      "next": "build"
+    },
+    {
+      "id": "build",
+      "name": "Build the transform",
+      "description": "apply the mapping and emit the target file",
+      "prompt": "Implement the transform exactly to the mapping table. Step 1: read the input with the correct delimiter/encoding. Step 2: apply each target column's rule in order (rename, cast, split/join, derive, constant); reshape (pivot/unpivot) if specified. Step 3: apply row filters and dedup; track rows-in vs rows-out and the reason for any drop. Step 4: write the output to data/output.csv (or the requested format/extension) with the target columns in the exact specified order and a header row. Step 5: ensure the output parses (valid CSV/JSON) and types are clean (no '1.0' where an int is expected). On a loop-back, fix only the named gaps. Call write_task_note with the path and rows-in/rows-out.",
+      "suggestedRole": "developer",
+      "advanceWhen": {
+        "file": "data/output.csv",
+        "minBytes": 1,
+        "sniff": "nonempty"
+      },
+      "gate": {
+        "at": "completion",
+        "checks": [
+          {
+            "kind": "minBytes",
+            "file": "data/output.csv",
+            "bytes": 120
+          },
+          {
+            "kind": "sniff",
+            "file": "data/output.csv",
+            "sniff": "data-table"
+          }
+        ],
+        "onReject": "build",
+        "maxAttempts": 4
+      },
+      "next": "evaluate"
+    },
+    {
+      "id": "evaluate",
+      "name": "Evaluate",
+      "description": "Grade the deliverable against every acceptance criterion. All pass → finish; any fail → loop back and fix the gap.",
+      "prompt": "Open data/output.csv and verify it matches the schema-scope contract. Step 1: confirm it parses in the target format and has EXACTLY the target columns in the specified order. Step 2: re-count the rows and reconcile against input — the count must equal input rows after the documented filter/dedup, with the delta explained. Step 3: pick 3 source rows and trace them through the mapping by hand to confirm each target field is correct end-to-end. Step 4: scan for type errors (e.g. '1.0' where an int is expected) and stray nulls the mapping didn't intend. Write PASS/FAIL per criterion; any FAIL loops back to build.\n\nThen route — this is the whole point of the loop:\n\n- **Every criterion PASSES →** call `advance_task_step({ ref, stepId: \"evaluate\", next: \"finish\" })`.\n- **Any criterion FAILS →** write the specific gaps to notes, then call `advance_task_step({ ref, stepId: \"evaluate\", next: \"build\" })` to loop back. The builder fixes exactly those gaps.\n\nNever route to `finish` while any criterion is unmet. The build phase's completion gate already blocked a grossly-incomplete deliverable; your job is the judgment an automated check cannot make (does it actually work, read well, look right). After ~3 unproductive loops, stop and report DONE_WITH_CONCERNS so the user can step in.",
+      "suggestedRole": "reviewer",
+      "consumes": [
+        {
+          "file": "data/output.csv"
+        }
+      ],
+      "next": "build"
+    },
+    {
+      "id": "finish",
+      "name": "Finish",
+      "description": "All acceptance criteria met. Stamp a short summary and report DONE.",
+      "prompt": "Every acceptance criterion passed. Write a one-paragraph DONE summary to task notes via `write_task_note`: what was built, the deliverable path(s), and a one-line confirmation that each criterion is met. Then report DONE.",
+      "suggestedRole": "developer",
+      "terminal": true
+    }
+  ],
+  "version": "1.2.2",
+  "releasedAt": "2026-08-14T00:00:00Z"
+}

package/data/craftbook-templates/cs/csv-transformer/versions/1.2.2/test.json

@@ -0,0 +1,102 @@
+{
+  "schemaVersion": 1,
+  "title": "Fixture-backed CSV normalization",
+  "objective": "Measure whether the csv-transformer craftbook locks a schema mapping and produces a transformed CSV from seeded source data.",
+  "tags": [
+    "html-page"
+  ],
+  "prompt": "In the `CSV Transformer Eval` project, read `data/source.csv` and write `data/output.csv`. Normalize to columns customer_id,full_name,total_cents,status. Trim whitespace, title-case names, convert dollar totals to integer cents, lowercase status, and de-duplicate by email keeping the latest row by updated_at. Keep the output sorted by customer_id.",
+  "setup": {
+    "projectName": "CSV Transformer Eval",
+    "about": "Self-contained eval project for csv-transformer. Source data is in workspace/data/source.csv and the deliverable is workspace/data/output.csv.",
+    "missionObjectives": "Transform the seeded messy CSV into a clean, sorted output CSV with normalized fields and duplicate handling.",
+    "files": [
+      {
+        "path": "data/source.csv",
+        "content": "customer_id,email,first,last,total_usd,status,updated_at\nC-001,ada@example.com, ada ,lane,12.99, ACTIVE ,2026-02-02\nC-002,bo@example.com,bo,ng,7.50,pending,2026-02-03\nC-001,ada@example.com,Ada,Lane,14.25,active,2026-02-05\nC-003,noor@example.com,noor,shah,25.00,Paused,2026-02-04\n"
+      }
+    ],
+    "worker": {
+      "name": "Nora",
+      "role": "Data Analyst"
+    }
+  },
+  "mocks": [],
+  "success": {
+    "summary": "data/output.csv has the expected normalized schema, deduped rows, integer cents, and lowercased statuses.",
+    "deliverables": [
+      {
+        "path": "data/output.csv",
+        "kind": "data-file",
+        "minBytes": 110,
+        "checks": [
+          {
+            "kind": "csvShape",
+            "file": "data/output.csv",
+            "exactColumns": [
+              "customer_id",
+              "full_name",
+              "total_cents",
+              "status"
+            ],
+            "minRows": 3,
+            "consistentColumns": true
+          },
+          {
+            "kind": "contains",
+            "file": "data/output.csv",
+            "pattern": "C-001,\"?Ada Lane\"?,1425,active",
+            "flags": "i"
+          },
+          {
+            "kind": "contains",
+            "file": "data/output.csv",
+            "pattern": "C-002,\"?Bo Ng\"?,750,pending",
+            "flags": "i"
+          },
+          {
+            "kind": "contains",
+            "file": "data/output.csv",
+            "pattern": "C-003,\"?Noor Shah\"?,2500,paused",
+            "flags": "i"
+          },
+          {
+            "kind": "notContains",
+            "file": "data/output.csv",
+            "pattern": "12\\.99|14\\.25|25\\.00",
+            "flags": "i"
+          }
+        ]
+      }
+    ]
+  },
+  "rubric": {
+    "artifact": {
+      "path": "data/output.csv",
+      "kind": "text"
+    },
+    "axes": [
+      {
+        "name": "structure",
+        "description": "Layout and hierarchy fit the content; sections are scannable and coherent."
+      },
+      {
+        "name": "grounding",
+        "description": "Seeded source content is used faithfully — no invented facts or dropped requirements."
+      },
+      {
+        "name": "interactivity",
+        "description": "Controls and states the page claims to offer actually work."
+      },
+      {
+        "name": "polish",
+        "description": "Styling is intentional and responsive rather than unstyled or broken."
+      }
+    ]
+  },
+  "qualityFocus": [
+    "schema mapping",
+    "dedupe rule",
+    "numeric normalization"
+  ]
+}

package/data/craftbook-templates/da/dashboard-spec/versions/1.0.1/craftbook.json

@@ -0,0 +1,107 @@
+{
+  "id": "dashboard-spec",
+  "name": "Dashboard Specification",
+  "description": "Specify a dashboard from the questions it must answer, before anyone builds it — the artifact that prevents a chart-salad dashboard nobody uses. Scopes the audience, the decisions, and the questions FIRST, then designs the dashboard layout: which widgets/charts answer which question, their metric definitions, filters, refresh, and visual hierarchy. The output is a build-ready spec, not the dashboard itself. Use this for planning a dashboard, a BI/reporting spec, a metrics dashboard design, deciding what charts a dashboard needs, or a dashboard requirements doc.\n\nA gallery craftbook generated from an archetype spec. It runs\n`phase → (per-phase gate) → … → evaluate → (loop) → finish`. Each build\nphase that produces a checkable artifact is followed by a **runtime\ngate-checkpoint** — the runtime verifies the artifact and routes with no\nmodel turn, looping back to redo the phase on a miss. The final `evaluate`\nstep holds a static deliverable gate plus a reviewer QA pass. What it adds\nover the generic `build-loop`: a specialist role per phase, a\ndomain-correct ordering, and a concrete per-phase quality bar.\n\nDeliverables marked \"artifact\" land in the project's artifacts drawer (`write_artifact` / `read_artifact`), not the shipped workspace — review output is not product source.\n\nPhases:\n\n1. Scope the questions (planner) — lock the audience, decisions, and questions the dashboard answers → gated on artifact `notes/question-scope.md` (markdown-notes)\n2. Design the dashboard (designer) — produce a build-ready layout: widgets, metrics, filters, hierarchy → gated on `dashboard-spec.md` (markdown-doc)\n\nThe gates never advance with an unmet criterion, and loop back to the\nowning phase to fix named gaps.\n",
+  "entryStepId": "question-scope",
+  "triggers": [
+    "spec a dashboard",
+    "plan a metrics dashboard",
+    "dashboard requirements",
+    "what charts should my dashboard have",
+    "bi dashboard design"
+  ],
+  "steps": [
+    {
+      "id": "question-scope",
+      "name": "Scope the questions",
+      "description": "lock the audience, decisions, and questions the dashboard answers",
+      "prompt": "Establish why the dashboard exists before designing widgets. Step 1: name the primary audience and the recurring decisions they make. Step 2: list the 5-10 specific questions the dashboard must answer for those decisions, ordered by importance. Step 3: for each question, name the metric(s) that answer it and define each metric precisely (numerator/denominator, time window). Step 4: identify the filters/segments users need (date range, team, region) and the refresh cadence. Step 5: write an acceptance-criteria checklist ('every widget traces to a listed question', 'each metric has a precise definition', 'top question is most prominent', 'required filters are listed', 'no orphan chart that answers no question'). Call write_task_note and write the audience + questions + metric definitions + checklist to notes/question-scope.md.\n\nThe deliverable `notes/question-scope.md` lands in the project's artifacts drawer — write it with `write_artifact` and read it back with `read_artifact`; the shipped workspace stays untouched.",
+      "suggestedRole": "planner",
+      "advanceWhen": {
+        "file": "notes/question-scope.md",
+        "minBytes": 1,
+        "sniff": "nonempty",
+        "artifact": true
+      },
+      "gate": {
+        "at": "completion",
+        "checks": [
+          {
+            "kind": "minBytes",
+            "file": "notes/question-scope.md",
+            "bytes": 120,
+            "artifact": true
+          },
+          {
+            "kind": "sniff",
+            "file": "notes/question-scope.md",
+            "sniff": "nonempty",
+            "artifact": true
+          }
+        ],
+        "onReject": "question-scope",
+        "maxAttempts": 3
+      },
+      "next": "design"
+    },
+    {
+      "id": "design",
+      "name": "Design the dashboard",
+      "description": "produce a build-ready layout: widgets, metrics, filters, hierarchy",
+      "prompt": "Design the dashboard as a build-ready specification (a spec, not the live dashboard). Step 1: choose the widget per question — KPI tile for a single headline number, line for trend, bar for comparison, table for detail — and justify each choice. Step 2: lay out the screen in priority order: a top KPI strip for the headline numbers, then trend and breakdown sections; describe the grid/wireframe in markdown. Step 3: for each widget specify title, chart type, the exact metric + its definition, dimensions/filters it responds to, and the data source/field. Step 4: specify global filters, default date range, and refresh cadence. Step 5: confirm every widget maps to a scoped question and there are no orphans. Write the full spec to dashboard-spec.md and call write_task_note with the widget count. On a loop-back, fix only the named gaps.",
+      "suggestedRole": "designer",
+      "advanceWhen": {
+        "file": "dashboard-spec.md",
+        "minBytes": 1,
+        "sniff": "nonempty"
+      },
+      "gate": {
+        "at": "completion",
+        "checks": [
+          {
+            "kind": "minBytes",
+            "file": "dashboard-spec.md",
+            "bytes": 500
+          }
+        ],
+        "scripts": [
+          {
+            "name": "checkContains",
+            "scope": "standard",
+            "inputs": {
+              "file": "dashboard-spec.md",
+              "pattern": "(?:^|\\n)#{1,3}\\s+\\S",
+              "flags": "i"
+            }
+          }
+        ],
+        "onReject": "design",
+        "maxAttempts": 4
+      },
+      "next": "evaluate"
+    },
+    {
+      "id": "evaluate",
+      "name": "Evaluate",
+      "description": "Grade the deliverable against every acceptance criterion. All pass → finish; any fail → loop back and fix the gap.",
+      "prompt": "Read dashboard-spec.md and check it against notes/question-scope.md. Verify: every widget traces explicitly to a listed question (and no orphan widget answers nothing); each widget names a chart type with a justification; each metric has a precise definition (numerator/denominator/window); the layout puts the top question most prominently; and required global filters, default date range, and refresh cadence are specified. Confirm a developer could build it without re-asking what a metric means. Write PASS/FAIL per criterion; any FAIL loops back to design.\n\nThen route — this is the whole point of the loop:\n\n- **Every criterion PASSES →** call `advance_task_step({ ref, stepId: \"evaluate\", next: \"finish\" })`.\n- **Any criterion FAILS →** write the specific gaps to notes, then call `advance_task_step({ ref, stepId: \"evaluate\", next: \"design\" })` to loop back. The builder fixes exactly those gaps.\n\nNever route to `finish` while any criterion is unmet. The build phase's completion gate already blocked a grossly-incomplete deliverable; your job is the judgment an automated check cannot make (does it actually work, read well, look right). After ~3 unproductive loops, stop and report DONE_WITH_CONCERNS so the user can step in.",
+      "suggestedRole": "reviewer",
+      "consumes": [
+        {
+          "file": "dashboard-spec.md"
+        }
+      ],
+      "next": "design"
+    },
+    {
+      "id": "finish",
+      "name": "Finish",
+      "description": "All acceptance criteria met. Stamp a short summary and report DONE.",
+      "prompt": "Every acceptance criterion passed. Write a one-paragraph DONE summary to task notes via `write_task_note`: what was built, the deliverable path(s), and a one-line confirmation that each criterion is met. Then report DONE.",
+      "suggestedRole": "developer",
+      "terminal": true
+    }
+  ],
+  "version": "1.0.1",
+  "releasedAt": "2026-08-14T00:00:00Z"
+}

package/data/craftbook-templates/da/dashboard-spec/versions/1.0.1/test.json

@@ -0,0 +1,114 @@
+{
+  "schemaVersion": 1,
+  "title": "Build-ready dashboard requirements spec",
+  "objective": "Measure whether the dashboard-spec craftbook scopes audience decisions and writes a build-ready metrics/dashboard spec instead of designing chart salad.",
+  "tags": [
+    "html-page"
+  ],
+  "prompt": "In the `Dashboard Spec Eval` project, read `source/dashboard-brief.md` and write `dashboard-spec.md` for a support operations dashboard. It must define audience, decisions, five questions, widgets/charts, metric definitions, filters, refresh cadence, visual hierarchy, and acceptance criteria. Do not build the dashboard.",
+  "setup": {
+    "projectName": "Dashboard Spec Eval",
+    "about": "Self-contained eval project for the dashboard-spec craftbook. Requirements are in workspace/source/dashboard-brief.md.",
+    "missionObjectives": "Produce a build-ready dashboard specification with audience, decisions, questions, widgets, metric definitions, filters, refresh cadence, hierarchy, and acceptance criteria.",
+    "files": [
+      {
+        "path": "source/dashboard-brief.md",
+        "content": "# Support Operations Dashboard Brief\n\nAudience: support director and queue leads.\nDecisions: staffing for Monday backlog, escalation coaching, and carrier status messaging.\nSignals available: ticket volume by queue, first response time, reopen rate, SLA breaches, carrier scan delay count, premium subscriber duplicate tickets, agent capacity, and weekly exception export status.\nRefresh: hourly during business hours, daily summary at 17:00.\nFilters: queue, region, carrier, customer segment, agent team, date range.\n"
+      }
+    ],
+    "worker": {
+      "name": "Lotte",
+      "role": "Planner"
+    }
+  },
+  "mocks": [],
+  "success": {
+    "summary": "workspace/dashboard-spec.md is a build-ready spec with audience, decisions, questions, widgets, metric definitions, filters, cadence, and hierarchy.",
+    "deliverables": [
+      {
+        "path": "dashboard-spec.md",
+        "kind": "markdown-report",
+        "minBytes": 2200,
+        "checks": [
+          {
+            "kind": "contains",
+            "file": "dashboard-spec.md",
+            "pattern": "Audience|support director|queue leads",
+            "flags": "i"
+          },
+          {
+            "kind": "contains",
+            "file": "dashboard-spec.md",
+            "pattern": "Decisions|staffing|escalation|carrier",
+            "flags": "i"
+          },
+          {
+            "kind": "contains",
+            "file": "dashboard-spec.md",
+            "pattern": "Questions|question",
+            "flags": "i"
+          },
+          {
+            "kind": "contains",
+            "file": "dashboard-spec.md",
+            "pattern": "Widget|chart|table|KPI",
+            "flags": "i"
+          },
+          {
+            "kind": "contains",
+            "file": "dashboard-spec.md",
+            "pattern": "metric definition|formula|SLA|reopen",
+            "flags": "i"
+          },
+          {
+            "kind": "contains",
+            "file": "dashboard-spec.md",
+            "pattern": "filter|queue|region|carrier|date range",
+            "flags": "i"
+          },
+          {
+            "kind": "contains",
+            "file": "dashboard-spec.md",
+            "pattern": "hourly|17:00|refresh",
+            "flags": "i"
+          },
+          {
+            "kind": "contains",
+            "file": "dashboard-spec.md",
+            "pattern": "acceptance criteria|hierarchy",
+            "flags": "i"
+          }
+        ]
+      }
+    ]
+  },
+  "rubric": {
+    "artifact": {
+      "path": "dashboard-spec.md",
+      "kind": "markdown"
+    },
+    "axes": [
+      {
+        "name": "structure",
+        "description": "Layout and hierarchy fit the content; sections are scannable and coherent."
+      },
+      {
+        "name": "grounding",
+        "description": "Seeded source content is used faithfully — no invented facts or dropped requirements."
+      },
+      {
+        "name": "interactivity",
+        "description": "Controls and states the page claims to offer actually work."
+      },
+      {
+        "name": "polish",
+        "description": "Styling is intentional and responsive rather than unstyled or broken."
+      }
+    ]
+  },
+  "qualityFocus": [
+    "question-first scope",
+    "metric definitions",
+    "build-ready spec"
+  ]
+}