Seasonal Maintenance Sweep
Run a season's home-care sweep against the project's systems roster: work out what care is due or overdue from each system's interval, plan the walkthrough in a sensible order with supplies, then log the completed care and flag what needs a professional — leaving the roster's due-tracking reset and a short sweep report. Locks the season's scope and the due list FIRST, then plans, then logs — so seasonal care becomes a checklist ritual instead of a vague worry. Use for spring/fall home maintenance, seasonal checkups, 'what does the house need', and pre-winter prep.
Steps
Entry step: scope. Each step names the specialist role it wants; the full working prompt is expandable.
- Scope the sweepplannerentry
season, due and overdue list from the roster intervals
Show working prompt
Define the sweep before planning it. Step 1: read systems.json from the workspace and determine the current season and which systems' care is due or overdue from each record's interval and last-done date, plus anything already flagged for attention. Step 2: with the owner, add any season-specific care the roster doesn't carry yet (gutters before the rains, hose bibs before frost) so the roster grows with each sweep. Step 3: split the list into owner-doable versus needs-a-professional. Step 4: write an acceptance-criteria checklist ('every due or overdue system is in the plan or explicitly deferred with a reason', 'each completed item gets its care logged so the due-tracking resets', 'professional items produce a booking note, not a vague intention', 'the report states what was done, deferred, and flagged'). Call write_task_note with the due counts + checklist and write notes/scope.md. - Plan the walkthroughplanner
ordered walkthrough with steps and supplies per system
Show working prompt
Plan the walkthrough from the locked list. Step 1: order the due items into a sensible route (outside before weather turns, top of the house down, wet work last). Step 2: for each item, write the short concrete care steps and what supplies or tools it needs, so the owner can walk the list without stopping to think. Step 3: batch the supply needs into one shopping line. Step 4: for the professional items, draft the booking note per trade — what is wrong or due, and the questions to ask. Step 5: write the ordered plan to notes/plan.md and call write_task_note with the item count and the supply list length.
- Log the sweepcopywriter
record completed care, flag findings, write the sweep report
Show working prompt
Close the sweep once the owner reports what got done. Step 1: for each completed item, log the maintenance through the project's house capabilities so the system's due-tracking resets. Step 2: for anything the walkthrough surfaced (a drip, a crack, a filter worse than expected), flag the system for attention with a note. Step 3: record what was deferred and why. Step 4: write the sweep report — done, deferred, flagged, and the professional bookings to make — short enough to read in a minute. Step 5: end with the next season's earliest due item so the ritual has a natural next date. Write report.md and call write_task_note with the done/deferred/flagged counts.
- Evaluatereviewer
Grade the deliverable against every acceptance criterion. All pass → finish; any fail → loop back and fix the gap.
Show working prompt
Open report.md and notes/plan.md and verify every criterion from notes/scope.md: every due or overdue system appears as done, deferred-with-reason, or flagged; every completed item's care was logged so its due-tracking reset (spot-check two against the roster); professional items carry a booking note; and the report states done/deferred/flagged plainly. Write PASS/FAIL per criterion; on any failure, name the gap and loop back to log. Then route — this is the whole point of the loop: - **Every criterion PASSES →** call `advance_task_step({ ref, stepId: "evaluate", next: "finish" })`. - **Any criterion FAILS →** write the specific gaps to notes, then call `advance_task_step({ ref, stepId: "evaluate", next: "log" })` to loop back. The builder fixes exactly those gaps. Never 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. - Finishdeveloper
All acceptance criteria met. Stamp a short summary and report DONE.
Show working 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.
Triggers
Phrases that suggest this craftbook to a crew.
- seasonal maintenance sweep
- what does the house need
- fall home checkup
- spring maintenance
- pre-winter prep
Source
View this craftbook on GitHub · MIT license