GURUKI LABS — SAMPLE SCENE BRIEF
Illustrative output · Prepared fictional workflow example
Concept and prototype design · Claude integration planned
Date: 2026-10-09

Purpose and current status
This prewritten mystery case explains a planned production workflow for Shorts and webtoon creators. It is not an AI-generated review, an actual creator approval, a customer project, or an evaluation result. The website buttons switch prepared examples; they do not save a project decision. This text file is the default sample, separate from the on-page selection.

Story and supplied context
A recurring mystery story: episode 3, an anonymous informant's notification.
R1 / Story rule: The informant's identity stays hidden until episode 5.
D1 / Prior decision: Keep the wristwatch visible. It is a clue for the next episode.
S3 / Current scene plan: The episode 3 phone shot shows the sender's name.
These are fictional supplied facts. What the characters already know, available assets, and other production conditions have not been supplied.

Review against the current intent
The sender name in S3 could expose the informant before R1 allows it. Review the name exposure, rather than treating every part of the scene as needing revision. Retain the visible wristwatch clue under D1.
Sources: R1 + S3 for the concern; D1 for the preserved detail.

Default example direction A / Hide the sender name
Change: Adjust the phone shot so the sender name is not visible.
Keep: Retain the visible wristwatch clue for the next episode.
Reason: Preserve the notification while avoiding an early identity reveal.
Sources: R1 + S3; D1.
Open question: Which shot adjustment fits the creator's available material? Production conditions still need to be supplied.
This is the default direction shown on the website and an example choice, not a real user's saved approval.

Alternative direction B / Use a reaction shot
Change: Show the character reacting to the notification and omit the phone screen.
Keep: Retain the visible wristwatch clue for the next episode.
Reason: Avoid showing the sender name while communicating the arrival of a notification.
Sources: R1 + S3; D1.
Open questions: Is suitable reaction material available? How will the visible wristwatch clue be retained in the framing or sequence? Confirm against the creator's production conditions.
The conceptual illustration is not evidence that the watch is visible in an actual shot.

Changed-intent counterexample / Keep the current plan
New explicit example intent: Reveal the informant's identity here in episode 3. This supersedes R1's episode-5 rule for this example.
Review: The sender name in S3 may remain visible. The supplied facts no longer require a sender-hiding revision.
Change: No sender-hiding revision is required under the new intent.
Keep: Retain the current phone shot and the visible wristwatch clue required by D1.
Reason: The new reveal timing permits the name exposure in S3.
Sources: New intent + S3; D1. R1 is superseded, not silently ignored.
Open question: Does this new reveal timing fit the rest of the story? Wider story context has not been supplied.
An adequate plan should be left alone. Information gaps should become focused questions.

Medium-specific production checks
Shorts: Review shot composition, when the sender information enters the frame, narration and timing, and whether the wristwatch clue remains readable.
Webtoon: Review panel composition, dialogue and information reveals, scrolling rhythm, and the panel or detail carrying the wristwatch clue.
These are applications of common story context to separate production plans, not an automatic conversion between media.

Before production
Confirm the creator's current reveal intent and relevant references. Check the chosen direction against available material and the visible watch requirement. Leave unprovided story facts, costs, timing, and feasibility open rather than inventing them. The creator makes the final production decision.

Planned Claude role and evaluation
The planned integration would ground review in supplied context, compare useful options, ask when information is unclear, leave adequate plans alone, and organize the creator's decisions into an editable scene brief.
Future evaluation will compare this workflow with creators' current tools and a general Claude chat, including context preparation and maintenance effort, useful decisions, reuse in production, and API cost and latency. No such evaluation results are claimed here.

한국어 요약
익명 제보자의 정체는 5화까지 숨긴다는 R1과 3화 휴대전화 화면에서 이름이 보이는 S3를 함께 검토하는 가상 예시다. 기본 예시 A는 이름을 가리고, 대안 B는 휴대전화 화면 대신 인물 반응을 보여준다. 두 안 모두 다음 화 단서인 손목시계를 유지하는 D1을 보존하되, 실제 소재와 화면 구성은 창작자가 확인해야 한다. 의도를 3화 정체 공개로 명시적으로 바꾸면 R1은 대체되며 현재 이름 노출 계획을 유지할 수 있다. 실제 AI 실행·고객 승인·제작 결과나 평가 성과를 주장하지 않는다.