원클릭으로
qrspi-design
Interactive design + structure/spec discussion — align on what we're building (QRSPI step 3 of 5)
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
메뉴
Interactive design + structure/spec discussion — align on what we're building (QRSPI step 3 of 5)
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Write the tactical implementation plan from all QRSPI artifacts (QRSPI step 5 of 5)
Generate objective research questions from a ticket (QRSPI step 1 of 5)
Objective codebase research from questions only — no ticket context (QRSPI step 2 of 5)
Create structure outline — vertical phases with test checkpoints (QRSPI step 4 of 5)
| name | qrspi_design |
| description | Interactive design + structure/spec discussion — align on what we're building (QRSPI step 3 of 5) |
You facilitate an interactive design discussion to build a shared understanding of what we're building. This is step 3 of the QRSPI workflow and produces the highest-leverage artifact: a ~200 line design document that captures every decision before a single line of code is written.
This is your chance to "brain dump" everything you think, everything you found, and everything you're unsure about — so the user can do surgery on your understanding before you go write 2,000 lines of code. Do not outsource the thinking. Force every decision to be explicit.
All QRSPI artifacts for a task live together in one directory:
thoughts/shared/tasks/ENG-XXXX-short-description/
├── ENG-XXXX-01-questions.md
├── ENG-XXXX-02-research.md
├── ENG-XXXX-03-design.md <-- you create this
├── ENG-XXXX-04-structure.md
└── ENG-XXXX-05-plan.md
You need two things:
ENG-XXXX-02-research.md in the task directory)If either is missing, ask for it. Read both FULLY before proceeding.
thoughts/shared/tasks/ENG-XXXX-description/ENG-XXXX-03-design.mdBased on the ticket and research, here's my understanding:
**Current State**: [what exists today, from research, with file:line refs]
**Desired End State**: [what the system looks like after we're done]
**Patterns I found** (confirm these are the RIGHT ones to follow):
- [Pattern A] — `file:line` — [brief description]
- [Pattern B] — `file:line` — [brief description]
**Patterns to AVOID** (ones I found that look wrong or outdated):
- [Anti-pattern] — `file:line` — [why I think we should avoid this]
**Design Questions** (need your input before I can proceed):
1. [Question about approach/tradeoff]
- A: [description + pros/cons]
- B: [description + pros/cons]
- Recommended: [your pick + why]
2. [Another question]
- A: ...
- B: ...
Wait for user responses. Ask follow-ups. Do NOT proceed until every question is answered and you have confirmed understanding.
Only write this AFTER all questions are resolved:
# Design: [Feature Name]
**Ticket**: [reference]
**Research**: [path to research doc]
**Date**: [today]
**Status**: draft
## Current State
[What exists today — from research, with file:line refs]
## Desired End State
[What the system looks like after implementation]
[How to verify we're done]
## Patterns to Follow
- [Pattern with file:line reference and brief description]
## Patterns to AVOID
- [Anti-pattern with explanation of why to avoid]
## Design Decisions
### 1. [Topic]
**Choice**: [what was decided]
**Reasoning**: [why, referencing user's input]
**Alternatives rejected**: [what was considered and why not]
### 2. [Topic]
...
## Constraints
- [Technical constraint from research]
- [Business constraint from ticket]
## Open Risks
- [Risk that implementation might surface]