Activity

Publication day is close, and the repair cafe has only scattered notes for its volunteer repair day. The work must make sense to owners of broken household items on a phone and lead naturally toward the choice to register an item. The producer organizes the assignment around human review and uses a simple side-by-side choice as the first point of recognition.

The planning sheet translates that search language into its underlying job: to evaluate unfamiliar image-making search language without implying endorsement. Success means a controlled test against one deliverable, not a large gallery of attractive drafts. It prevents a broad technology discussion from replacing the work needed for the volunteer repair day.

The first writing pass produces four message routes, not a pile of final captions. Four routes expose a simple side-by-side choice, show a three-step method, compare responsible choices, or document one decision at the repair cafe. Prompt it. Shape it. Make it yours. are scored for fit with human review, evidence, visual clarity, and whether they lead naturally to register an item.

The center of the one-page brief is simple: help owners of broken household items understand the volunteer repair day and decide whether to register an item. Voice guidance says resourceful and calm, while the facts column holds timing, price, eligibility, fulfilment, and contact details. Guarantees and implied affiliations stay out. Production fields cover exact spelling and case, included and avoided words, sensitive claims, source permissions, a 4:5 master crop, mobile margins, a 30-second limit, formats, owners, and publishing time. Three labels matter: confirmed, verify before use, and illustrative only.

The approved route expands into one explanation, a caption, a two-line hook, a five-card carousel, a voiceover, and three headline options. The five cards move from audience problem to required facts, illustrative choice, human check, and the action to register an item. If a fact is absent, the draft carries an editor query and stops short of publication wording. The copy earns space by helping the reader act, judge, compare, or avoid a known failure.

A vertical script capped at 30 seconds shows one decision, from loose notes about the volunteer repair day to a reviewed asset. For three seconds, voiceover names a simple side-by-side choice, screen copy stays under five words, and the camera shows real planning notes after private details are removed. Shot two collects inputs for four seconds; shot three shows the illustrative choice for six; shot four pauses on a human correction for five; shot five uses the remaining time for the approved next step. Each shot row keeps narration, screen copy, camera instruction, timing, evidence status, and reviewer distinct.

For the image route, the single job is to turn three selection criteria into a visual sequence. Prompt fields cover the volunteer repair day, spatial relationship, dominant focus, plain setting, the palette deep green, warm gray, and off-white, believable side light, and 4:5 output. Upper and lower safe zones remain quiet for human-added type, while the central action stays readable at phone width. The sample labels ‘Made Nearby’ and ‘First Things First’ are illustrative examples only; no testing or performance is implied.

Specific model failures in this case include invented features and unsupported performance claims. A system may also change the subject between frames, deform text or interfaces, miss cultural meaning, imitate a familiar mark, or describe an example as a real result. A second reviewer compares each output with the source, inspects artwork at full size, and reads the script without prompt history.

The schedule also reserves enough time for correction, source checks, and another careful final reading before any finished asset is queued. The minimum viable set for the volunteer repair day contains a master argument, a flexible visual frame, related caption copy, and a timed demonstration of the same decision. The owner reviews that chain before variants are made, because a weak assumption copied into twelve exports becomes harder to locate and more expensive to repair. The production board records the asset, audience purpose, supporting fact, destination format, approval state, reviewer, and how it helps people register an item. The learning note defines one observable response, records date range and exposure, and avoids crediting a creative change for movement that the available evidence cannot explain. The result informs the next human review decision but is not presented as universal proof, a guaranteed lift, or a customer testimonial. Version control is deliberately plain: drafts receive descriptive names, approval is written down, and obsolete exports move out of the active folder so nobody publishes an attractive but rejected file. Before scheduling, the producer checks the actual derivatives in their destination previews for compression, crop, caption placement, links, and any change to the intended reading order.

Core facts stay fixed, while opening, order, pace, composition, and interaction are rebuilt for each destination. The written channel preserves the decision process, the image channel leads with the clearest frame, and the carousel reveals one step at a time. Vertical video presents a simple side-by-side choice immediately and remains phone-safe. Long-form keeps the full comparison, while a community edit requests feedback before people register an item.

Human approval starts by matching every claim to confirmed inputs, checking current external rules against reliable sources, recording the access date, and testing audience fit and voice. The visual reviewer checks 4:5 derivatives, mobile crop, contrast, hierarchy, spelling, doubled characters, icons, anatomy, edges, shadows, duplicate objects, interface structure, and accidental marks. A cold reviewer watches for changing objects, warped text, caption errors, unsafe margins, mispronunciation, uneven pace or audio, weak boundary frames, and lost meaning without sound. Final approval is recorded when the formats hold the same facts, maintain the stated resourceful and calm tone, and make the next action appropriate for owners of broken household items.

Handoff keeps approved inputs, copy route, working text, layered design, captions, shot table, source notes, and checked exports under distinct names. With one source of truth, the producer can change a detail and identify every affected placement. The finished set stays modest: one approved idea, channel-specific expressions, and a documented human decision about the volunteer repair day.