@bendyline/gilde 0.1.38 → 0.1.39

package/data/craftbook-templates/me/meeting-notes-to-actions/versions/1.0.2/craftbook.json

@@ -0,0 +1,119 @@
+{
+  "id": "meeting-notes-to-actions",
+  "name": "Meeting Notes to Action Items",
+  "description": "Convert raw meeting notes or a transcript into a clean, assigned action-item list with owners, due dates, and the decisions that drove them. Extracts the actions and decisions FIRST (every commitment, ask, and decision lifted verbatim-ish into a structured list with source quotes), then a planner assigns owners and due dates and prioritizes — separating extraction from assignment so nothing said in the meeting is lost before triage. The extract-then-assign ordering is what stops a small model from silently dropping commitments or inventing owners. Use for meeting follow-ups, action-item extraction, turning transcripts into to-dos, decision logs, and post-meeting recaps.\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. Extract actions & decisions (developer) — lift every commitment, ask, and decision with source quotes → gated on artifact `{{workPath}}/extract.md` (markdown-notes)\n2. Assign & prioritize (planner) — set owners, due dates, and priority into the final list → gated on `actions.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": "extract",
+  "triggers": [
+    "meeting notes to action items",
+    "extract action items",
+    "meeting follow-up",
+    "turn transcript into tasks",
+    "post-meeting recap"
+  ],
+  "steps": [
+    {
+      "id": "extract",
+      "name": "Extract actions & decisions",
+      "description": "lift every commitment, ask, and decision with source quotes",
+      "prompt": "Read the meeting {{workPath}}/transcript end to end and extract — do not yet assign or prioritize. (1) Pull EVERY action item: anything phrased as a commitment, follow-up, ask, or 'someone should…'. For each, capture the action verbatim-ish, any owner/date mentioned, and the source quote it came from. (2) Pull every DECISION (a choice the group landed on) and every OPEN QUESTION left unresolved, each with its source quote. (3) Lock the output schema for the next phase: action | owner | due | priority | source-quote. Be exhaustive — recall matters more than precision here; a borderline item is kept, not dropped. Write an acceptance-criteria checklist ('every commitment in the notes is captured', 'each action cites a source quote', 'decisions and open questions separated from actions', 'no owner/date invented beyond what the notes state') plus the raw extracted lists to {{workPath}}/extract.md via write_task_note AND the produces path.\n\nThe deliverable `{{workPath}}/extract.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": "developer",
+      "advanceWhen": {
+        "file": "{{workPath}}/extract.md",
+        "minBytes": 1,
+        "sniff": "nonempty",
+        "artifact": true
+      },
+      "gate": {
+        "at": "completion",
+        "checks": [
+          {
+            "kind": "minBytes",
+            "file": "{{workPath}}/extract.md",
+            "bytes": 120,
+            "artifact": true
+          },
+          {
+            "kind": "sniff",
+            "file": "{{workPath}}/extract.md",
+            "sniff": "nonempty",
+            "artifact": true
+          }
+        ],
+        "onReject": "extract",
+        "maxAttempts": 3
+      },
+      "next": "assign"
+    },
+    {
+      "id": "assign",
+      "name": "Assign & prioritize",
+      "description": "set owners, due dates, and priority into the final list",
+      "prompt": "Turn the extracted items into actions.md, the deliverable. (1) Render a table: Action | Owner | Due | Priority | Source. (2) Assign the owner stated in the notes; where the notes name no owner, write 'UNASSIGNED — [proposed: name]' rather than guessing silently. (3) Set due dates from the notes; where none exists, leave 'TBD' (do not fabricate dates). (4) Assign priority (P1/P2/P3) using a stated, simple rule (P1 = explicit deadline or blocker, else P2, nice-to-haves P3) and note the rule at the top. (5) Below the table add a Decisions section and an Open questions section from the extraction. Every extracted action must appear — none dropped in triage. On a loop-back fix only the flagged rows. Write the path and a count (actions, unassigned, decisions) to notes via write_task_note.",
+      "suggestedRole": "planner",
+      "advanceWhen": {
+        "file": "actions.md",
+        "minBytes": 1,
+        "sniff": "nonempty"
+      },
+      "gate": {
+        "at": "completion",
+        "checks": [
+          {
+            "kind": "minBytes",
+            "file": "actions.md",
+            "bytes": 500
+          }
+        ],
+        "scripts": [
+          {
+            "name": "checkContains",
+            "scope": "standard",
+            "inputs": {
+              "file": "actions.md",
+              "pattern": "(?:^|\\n)#{1,3}\\s+\\S",
+              "flags": "i"
+            }
+          }
+        ],
+        "onReject": "assign",
+        "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": "QA actions.md against the locked criteria. (1) Cross-check every action in {{workPath}}/extract.md against the final table — all present? FAIL on any dropped commitment. (2) Confirm each row has a source reference and that no owner or due date was invented beyond what the notes state (unstated owners marked UNASSIGNED, unstated dates TBD). (3) Confirm Decisions and Open-questions sections exist and match the extraction. (4) Confirm the priority rule is stated and applied consistently on 2-3 rows. (5) Re-scan the original notes for any commitment the extraction missed. Write PASS/FAIL per criterion with the specific items.\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: \"assign\" })` 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": "actions.md"
+        }
+      ],
+      "next": "assign"
+    },
+    {
+      "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": "1.0.2",
+  "releasedAt": "2026-08-21T00:00:00Z",
+  "paramSchema": {
+    "type": "object",
+    "properties": {
+      "workPath": {
+        "type": "string",
+        "title": "Working folder",
+        "description": "Per-task working folder in the artifacts drawer. Defaults to this task's own folder so runs never collide; override with a stable name when you deliberately want runs to share files.",
+        "default": "{{task.dir}}"
+      }
+    }
+  },
+  "minGezelVersion": "1.26233"
+}

package/data/craftbook-templates/me/meeting-notes-to-actions/versions/1.0.2/test.json

@@ -0,0 +1,113 @@
+{
+  "schemaVersion": 1,
+  "title": "Meeting Notes to Action Items smoke eval",
+  "objective": "Self-contained smoke eval for the Meeting Notes to Action Items craftbook using the corpus generic harness.",
+  "tags": [
+    "corpus"
+  ],
+  "prompt": "We need a meeting notes to action items 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": "Meeting Notes to Action Items Eval",
+    "about": "Self-contained eval project for meeting-notes-to-actions. Seeded inputs are under workspace/source or workspace/fixtures; final deliverable is workspace/report.md.",
+    "missionObjectives": "Use the Meeting Notes to Action Items 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": "# Meeting Notes to Action Items 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: meeting-notes-to-actions - Meeting Notes to Action Items.\n"
+      }
+    ],
+    "worker": {
+      "name": "Jules",
+      "role": "Writer"
+    },
+    "craftbookParams": {
+      "workPath": "tasks/eval"
+    }
+  },
+  "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/me/meeting-prep-brief/versions/1.0.2/craftbook.json

@@ -0,0 +1,119 @@
+{
+  "id": "meeting-prep-brief",
+  "name": "Meeting Prep Brief",
+  "description": "Walk into the meeting already warm: who's in the room, what was promised last time, what you want out of this one, and the two questions worth asking — one page, read in the elevator. Locks the target meeting and the one-page criteria FIRST, then compiles from the calendar and the prior notes, so every claim in the brief traces to something real instead of memory. Use for meeting prep, 'prep me for the meeting', pre-meeting briefs, vendor and client syncs, and walking in warm.\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 meeting (planner) — target meeting, attendees from the record, history gathered → gated on artifact `{{workPath}}/scope.md` (markdown-notes)\n2. Write the brief (copywriter) — one page: room, goal, last time, agenda, the questions → gated on artifact `{{workPath}}/report.md` (markdown-report)\n\nThe gates never advance with an unmet criterion, and loop back to the\nowning phase to fix named gaps.\n",
+  "entryStepId": "scope",
+  "triggers": [
+    "prep me for the meeting",
+    "meeting prep brief",
+    "get me ready for the meeting",
+    "brief me before the meeting",
+    "what do i need for this meeting"
+  ],
+  "steps": [
+    {
+      "id": "scope",
+      "name": "Scope the meeting",
+      "description": "target meeting, attendees from the record, history gathered",
+      "prompt": "Pick the meeting and gather everything before writing a word. Step 1: identify the target meeting — the one the owner names, or the next upcoming one worth prep — from calendar.json in the workspace or through the project's calendar capabilities when the calendar is wired as a live service, and pull its details: when, where, and exactly who is in the room. Step 2: gather the history — prior notes for this meeting or these people under {{workPath}}/, and the open commitments made last time (who promised what, and whether it happened). Step 3: agree with the owner what they want out of this meeting — the one outcome that makes it a success. Step 4: write an acceptance-criteria checklist ('attendees listed from the calendar record, not from memory', 'every last-time commitment appears with its current status', 'the goal is stated in one sentence', 'at least two questions worth asking, ranked', 'one page — readable in the elevator'). Call write_task_note with the meeting + attendee count + checklist and write {{workPath}}/scope.md.\n\nThe deliverable `{{workPath}}/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": "{{workPath}}/scope.md",
+        "minBytes": 1,
+        "sniff": "nonempty",
+        "artifact": true
+      },
+      "gate": {
+        "at": "completion",
+        "checks": [
+          {
+            "kind": "minBytes",
+            "file": "{{workPath}}/scope.md",
+            "bytes": 120,
+            "artifact": true
+          },
+          {
+            "kind": "sniff",
+            "file": "{{workPath}}/scope.md",
+            "sniff": "nonempty",
+            "artifact": true
+          }
+        ],
+        "onReject": "scope",
+        "maxAttempts": 3
+      },
+      "next": "brief"
+    },
+    {
+      "id": "brief",
+      "name": "Write the brief",
+      "description": "one page: room, goal, last time, agenda, the questions",
+      "prompt": "Write the one-page brief from what you gathered. Step 1: header — the meeting, when, and who is in the room, straight from the calendar details. Step 2: the goal — what the owner wants out of this one, in one sentence near the top. Step 3: last time — each commitment from the prior notes with its honest status: done, in motion, or silently dropped (the dropped ones are usually the real meeting). Step 4: the agenda as you would run it — short, ordered, loosely timed. Step 5: the questions — at least two worth asking, ranked, phrased ready to say out loud. Keep every claim traceable to the calendar or the notes; where the record is silent, say so rather than invent. One page. Write {{workPath}}/report.md and call write_task_note with the goal and the top question.\n\nThe deliverable `{{workPath}}/report.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": "copywriter",
+      "advanceWhen": {
+        "file": "{{workPath}}/report.md",
+        "minBytes": 1,
+        "sniff": "nonempty",
+        "artifact": true
+      },
+      "gate": {
+        "at": "completion",
+        "checks": [
+          {
+            "kind": "minBytes",
+            "file": "{{workPath}}/report.md",
+            "bytes": 1000,
+            "artifact": true
+          },
+          {
+            "kind": "contains",
+            "file": "{{workPath}}/report.md",
+            "pattern": "(?:^|\\n)#{1,3}\\s+\\S",
+            "flags": "i",
+            "label": "at least one markdown heading",
+            "artifact": true
+          }
+        ],
+        "onReject": "brief",
+        "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 {{workPath}}/report.md, the prior notes under {{workPath}}/, and the calendar details and verify every criterion from {{workPath}}/scope.md: every attendee in the brief matches the calendar record, every commitment listed traces to the prior notes with an honest status, the goal is one sentence, at least two ranked questions are present and phrased to ask aloud, nothing in the brief is invented — every claim has a source in the notes or the calendar — and it fits one page. Write PASS/FAIL per criterion; on any failure, name the gap and loop back to brief.\n\nThe deliverable lives in the project's artifacts drawer — open `{{workPath}}/report.md` with `read_artifact`, not `read_file`.\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: \"brief\" })` 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": "{{workPath}}/report.md",
+          "artifact": true
+        }
+      ],
+      "next": "brief"
+    },
+    {
+      "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": "1.0.2",
+  "releasedAt": "2026-08-21T00:00:00Z",
+  "paramSchema": {
+    "type": "object",
+    "properties": {
+      "workPath": {
+        "type": "string",
+        "title": "Working folder",
+        "description": "Per-task working folder in the artifacts drawer. Defaults to this task's own folder so runs never collide; override with a stable name when you deliberately want runs to share files.",
+        "default": "{{task.dir}}"
+      }
+    }
+  },
+  "minGezelVersion": "1.26233"
+}

package/data/craftbook-templates/me/meeting-prep-brief/versions/1.0.2/test.json

@@ -0,0 +1,187 @@
+{
+  "schemaVersion": 1,
+  "title": "Meeting Prep Brief smoke eval",
+  "objective": "Task-specific eval for the Meeting Prep Brief craftbook: the vendor sync's attendees come from a live fake calendar service, the prior meeting's notes carry three commitments (one silently dropped), and tasks/eval/report.md is a one-page brief that surfaces all of it plus ranked questions.",
+  "tags": [
+    "external"
+  ],
+  "prompt": "ive got the vendor sync thursday — can you prep me a brief? last meetings notes are in tasks/eval/ and the event details come from the fake calender service, theres a probe script wired for it",
+  "setup": {
+    "projectName": "Calendar Desk — Meeting Prep Eval",
+    "about": "Self-contained eval project for meeting-prep-brief. The calendar is wired as a live fake service — endpoints in mocks/services.md, and the ready-made mock-probe project script fetches the upcoming events and the target event's detail. Prior meeting notes are seeded under tasks/eval/; the deliverable is tasks/eval/report.md.",
+    "missionObjectives": "Use the Meeting Prep Brief craftbook/template, pull the vendor sync's details (attendees included) from the fake calendar service through the provided probe, ground every claim in the prior notes, and write tasks/eval/report.md. No real services, credentials, or live calendars — only the wired mock.",
+    "files": [
+      {
+        "path": "tasks/eval/vendor-sync-2026-06.md",
+        "content": "# Vendor sync - Northwind Systems - 2026-06-18\n\nAttending: Dana Whitfield, Marcus Lee (Northwind); Priya Raman, me.\n\n## What happened\n- Walked the integration scope; Northwind demoed the sandbox tenant.\n- Pricing came up again - they hinted the per-seat tier could move if we commit by Q3.\n\n## Commitments\n- We send the revised pricing tiers by July 1. (sent June 30 - done)\n- Dana sends the data-migration timeline by June 25. (never arrived - still open)\n- Both sides name a launch owner. (ours is Priya, theirs is Marcus - done)\n\n## Parking lot\n- Sandbox rate limits looked tight for our nightly sync - nobody owned this.\n"
+      }
+    ],
+    "worker": {
+      "name": "Meike",
+      "role": "Planner"
+    },
+    "craftbookParams": {
+      "workPath": "tasks/eval"
+    }
+  },
+  "mocks": [
+    {
+      "kind": "http",
+      "id": "calendar",
+      "description": "Fake calendar service for this eval. GET /api/events/upcoming lists the next events; GET /api/events/EVT-204 returns the vendor sync's full detail including attendees.",
+      "credential": {
+        "name": "mock.calendar"
+      },
+      "routes": [
+        {
+          "method": "GET",
+          "path": "/api/events/upcoming",
+          "body": {
+            "events": [
+              {
+                "id": "EVT-204",
+                "title": "Vendor sync - Northwind Systems",
+                "start": "2026-07-23T14:00",
+                "end": "2026-07-23T15:00",
+                "location": "video"
+              },
+              {
+                "id": "EVT-205",
+                "title": "Design review",
+                "start": "2026-07-24T09:30",
+                "end": "2026-07-24T10:30",
+                "location": "studio"
+              }
+            ]
+          }
+        },
+        {
+          "method": "GET",
+          "path": "/api/events/EVT-204",
+          "body": {
+            "id": "EVT-204",
+            "title": "Vendor sync - Northwind Systems",
+            "start": "2026-07-23T14:00",
+            "end": "2026-07-23T15:00",
+            "location": "video",
+            "attendees": [
+              "Dana Whitfield (Northwind, account lead)",
+              "Marcus Lee (Northwind, migration engineer)",
+              "Priya Raman (our launch owner)",
+              "you"
+            ],
+            "agenda": null
+          }
+        }
+      ]
+    },
+    {
+      "kind": "cli",
+      "id": "probe",
+      "description": "Provenance-trusted project script `mock-probe` that lists the upcoming events and fetches the vendor sync's detail from the fake calendar through http.authed.",
+      "shim": {
+        "path": "scripts/mock-probe.ts",
+        "content": "import { defineScript, gezel } from '@bendyline/gezel-sdk';\n\nexport const meta = defineScript({\n  name: 'mock-probe',\n  description:\n    'Probe the live fake calendar service for this eval: lists the upcoming events and fetches the full detail (attendees included) of the first one. Outputs the raw JSON responses as evidence for the brief.',\n  inputs: {},\n  outputs: {\n    upcoming: { type: 'string', description: 'Raw JSON body from the upcoming-events listing.' },\n    detail: { type: 'string', description: 'Raw JSON body of the first upcoming event, attendees included.' },\n  },\n  requires: ['workspace.read', 'network', 'credential:mock.calendar'],\n});\n\nconst services = JSON.parse(await gezel.fs.read('mocks/services.json'));\nconst base = (id) => {\n  const service = services.find((entry) => entry.id === id);\n  if (!service) throw new Error(`mock service \"${id}\" is not listed in mocks/services.json`);\n  return service.baseUrl;\n};\n\nconst upcoming = await gezel.http.authed(`${base('calendar')}/api/events/upcoming`, {\n  credential: 'mock.calendar',\n});\nif (!upcoming.ok) throw new Error(`upcoming listing failed: ${upcoming.status} ${upcoming.body}`);\n\nconst first = JSON.parse(upcoming.body).events[0];\nconst detail = await gezel.http.authed(`${base('calendar')}/api/events/${first.id}`, {\n  credential: 'mock.calendar',\n});\nif (!detail.ok) throw new Error(`event detail failed: ${detail.status} ${detail.body}`);\n\ngezel.output({ upcoming: upcoming.body, detail: detail.body });\n"
+      }
+    }
+  ],
+  "success": {
+    "summary": "tasks/eval/report.md is a one-page brief grounded in the live fake calendar and the prior notes: attendees match the calendar record, the never-delivered migration timeline resurfaces with its status, and at least two ranked questions are ready to ask aloud.",
+    "deliverables": [
+      {
+        "path": "tasks/eval/report.md",
+        "kind": "markdown-report",
+        "minBytes": 700,
+        "checks": [
+          {
+            "kind": "contains",
+            "file": "tasks/eval/report.md",
+            "pattern": "(?:^|\\n)#{1,3}\\s+\\S",
+            "flags": "i"
+          },
+          {
+            "kind": "contains",
+            "file": "tasks/eval/report.md",
+            "pattern": "Dana",
+            "flags": "i",
+            "label": "attendees come from the live calendar record"
+          },
+          {
+            "kind": "contains",
+            "file": "tasks/eval/report.md",
+            "pattern": "Marcus",
+            "flags": "i",
+            "label": "the second Northwind attendee is named"
+          },
+          {
+            "kind": "contains",
+            "file": "tasks/eval/report.md",
+            "pattern": "commit|promis|agreed|owe",
+            "flags": "i",
+            "label": "last-time commitments appear with status"
+          },
+          {
+            "kind": "contains",
+            "file": "tasks/eval/report.md",
+            "pattern": "migration|timeline",
+            "flags": "i",
+            "label": "the silently-dropped migration timeline resurfaces"
+          },
+          {
+            "kind": "contains",
+            "file": "tasks/eval/report.md",
+            "pattern": "\\?",
+            "label": "the brief asks real questions"
+          },
+          {
+            "kind": "judge",
+            "file": "tasks/eval/report.md",
+            "rubric": "The brief reads in under two minutes, and every claim traces to the prior meeting notes or the calendar detail — nothing invented. The never-delivered data-migration timeline is surfaced as an open item rather than forgotten.",
+            "sourceFiles": [
+              "tasks/eval/vendor-sync-2026-06.md"
+            ],
+            "label": "grounded and elevator-readable"
+          }
+        ]
+      }
+    ],
+    "mocks": [
+      {
+        "service": "calendar",
+        "minRequests": 1,
+        "requiredPaths": [
+          "^/api/events"
+        ]
+      }
+    ]
+  },
+  "rubric": {
+    "artifact": {
+      "path": "tasks/eval/report.md",
+      "kind": "markdown"
+    },
+    "axes": [
+      {
+        "name": "grounding",
+        "description": "Every attendee, commitment, and claim traces to the calendar record or the prior notes — nothing from memory or imagination."
+      },
+      {
+        "name": "brevity",
+        "description": "One page, elevator-readable; the goal is one sentence near the top."
+      },
+      {
+        "name": "question-quality",
+        "description": "At least two ranked questions, phrased to say out loud, aimed at what actually matters."
+      },
+      {
+        "name": "follow-through",
+        "description": "The dropped commitment (the migration timeline that never arrived) is surfaced as the real business of the meeting."
+      }
+    ]
+  },
+  "qualityFocus": [
+    "calendar grounding",
+    "commitment tracking",
+    "one-page discipline"
+  ]
+}