The story centers on Internet product teams teams who capture review conclusions, issues, and edits in one list.
Overview
Background
Core collaborators include Designer, Product manager, Frontend lead, each adding context in a fast-moving thread.
Without a structured capture, decisions blur and follow-through slows.
Intervention
When the thread hits a decision point, the team triggers “@Holycrab_AI summarize the design review and output a change list”. Holycrab AI then Design files, review comments, interaction constraints, Cluster disputes -> log decisions -> split edits, Review decisions + edit list + decision log.
The distilled output stays tied to the original context for quick review.
- Input Design files, review comments, interaction constraints
- Process Cluster disputes -> log decisions -> split edits
- Output Review decisions + edit list + decision log
Deliverables & Impact
Deliverables include Review decisions, Edit list, Decision log, giving the team a clear handoff.
Impact shows up as Wrap-up time: 60 min -> 10 min; Missed issues: Lower.
- Review decisions Keep vs change conclusions
- Edit list Edits with owners
- Decision log Key trade-offs and reasons
Example
Prompt: @Holycrab_AI summarize the design review and output a change list
Decision: keep the two-column layout
Edits: button contrast and states
Owner: design update in 2 days
