New Product - Deep Research & Architecture Design
Given documentation or a product description, perform deep research and cross-reference technologies to determine the optimal architecture for building a new application.
Your Task
You are orchestrating a 6-stage deep research workflow to design product architecture. Run the stages in order, use 4 parallel agents for the deep-research stages, and present a final comprehensive review at the end.
All document and agent-prompt templates live in this skill's references/templates/ — read the listed template at each stage; do not improvise document structures.
Stage 0: Initialize Product Workspace
- Parse the input argument: if it is a file path (contains
/ or ends in .md, .txt, etc.), read the documentation; otherwise treat it as a product description.
- Extract a short kebab-case product name (from the doc's project/product title, or derived from the description, e.g. "real-time chat app" →
realtime-chat). Store it for later stages.
mkdir -p /job-queue/product-{name}/research-notes- Display: "🚀 Stage 0/6: Workspace created at /job-queue/product-{name}/"
Stage 1: Big Idea & High-Level Research
- Analyze input. From documentation: extract features, user flows, requirements, technical hints/constraints. From a description: core functionality, target users, key features, scale expectations.
- Ask 3 initial clarifying questions via
AskUserQuestion — Product Vision, Scale & Audience, Deployment Context. Exact question/option text: references/templates/stage1-big-idea.md.
- High-level technology research via
WebSearch: "[product type] architecture patterns 2026", "[product type] technology stack comparison 2026", "[product type] best practices 2026", "[scale level] architecture considerations 2026". WebFetch the top 2-3 articles per search; extract common patterns, recommended technologies, pitfalls. Save notes to /job-queue/product-{name}/research-notes/stage1-research.md.
- Write
/job-queue/product-{name}/big-idea.md using the content structure in references/templates/stage1-big-idea.md (overview, value proposition, architecture approach, frontend/backend/database/deployment options with pros/cons, key challenges, next steps, research sources).
- Display: "✅ Stage 1/6: Big idea generated at product-{name}/big-idea.md"
Stages 2-5: Parallel Deep Research for Architectural Documents
Launch all 4 research agents at once via the Task tool with subagent_type="general-purpose". Each agent researches one architectural document independently.
| Stage |
Task description |
Output document |
Prompt template |
| 2 |
"Runtime execution research" — how the system executes work |
runtime-execution.md |
references/templates/stage2-runtime-execution.md |
| 3 |
"Abstraction layer research" — how user intent becomes executable logic |
abstraction-layer.md |
references/templates/stage3-abstraction-layer.md |
| 4 |
"Integration layer research" — how the system connects to external resources |
integration-layer.md |
references/templates/stage4-integration-layer.md |
| 5 |
"Output rendering research" — how results are delivered to consumers |
output-rendering.md |
references/templates/stage5-output-rendering.md |
Build each agent prompt from its template:
- Read the stage's template plus
references/templates/common-closing-sections.md — the single deduplicated tail (Technology Comparison / Implementation Considerations / Research Sources) shared by all four documents; insert it at the template's [COMMON CLOSING SECTIONS] marker with that stage's variants.
- Substitute
[Product Name] and {name}, and paste the full big-idea.md content into the CONTEXT block (every agent needs it for cross-referencing).
Each template already instructs the agent to WebSearch/WebFetch, ask its own AskUserQuestion clarifications, and write both its document and raw notes under /job-queue/product-{name}/.
Wait for all 4 agents to complete, then display "✅ Stage N/6: [topic] research complete" for stages 2-5.
Stage 6: Final Review & Presentation
- Read all 5 generated documents (big-idea.md + the 4 architecture docs).
- Write
/job-queue/product-{name}/ARCHITECTURE-SUMMARY.md (stack, style, key decisions, trade-offs, risks, next steps) and print the completion banner — both templates in references/templates/stage6-summary-and-review.md.
- Ask for feedback via
AskUserQuestion (exact text in the same template): Approve, or Revise runtime execution / abstraction layer / integration layer / output rendering.
- Approve → display "✅ Architecture approved! Ready for implementation."
- Revise → re-run that stage's research agent with the feedback, regenerate the document, re-present for approval.
Error Handling
On any stage failure: display which stage failed, list completed vs failed stages, show the error, note partial work saved at /job-queue/product-{name}/, and give recovery instructions (resume with /new-product [input] or continue manually). Example failure display: references/templates/stage6-summary-and-review.md.
Notes
- Tools: WebSearch + WebFetch (deep research — use extensively), Task (parallel agents), AskUserQuestion (iterative clarification), Read/Write (documents).
- Ask clarifying questions during each stage, not all upfront; each agent reads big-idea.md for context. Documents must be comprehensive, well-structured, actionable.
- Duration ~20-40 min: Stage 0 seconds; Stage 1 ~5-10 min; Stages 2-5 ~15-25 min (parallel); Stage 6 ~2-3 min.
- Done = 6 documents in
/job-queue/product-{name}/ (big-idea, 4 architecture docs, ARCHITECTURE-SUMMARY), trade-offs and risks identified, user approval obtained.
1---2name: new-product3description: Deep research and architecture design for new product development4---56# New Product - Deep Research & Architecture Design78Given documentation or a product description, perform deep research and cross-reference technologies to determine the optimal architecture for building a new application.910## Your Task1112You are orchestrating a 6-stage deep research workflow to design product architecture. Run the stages in order, use 4 parallel agents for the deep-research stages, and present a final comprehensive review at the end.1314All document and agent-prompt templates live in this skill's `references/templates/` — read the listed template at each stage; do not improvise document structures.1516---1718## Stage 0: Initialize Product Workspace1920- Parse the input argument: if it is a file path (contains `/` or ends in `.md`, `.txt`, etc.), read the documentation; otherwise treat it as a product description.21- Extract a short kebab-case product name (from the doc's project/product title, or derived from the description, e.g. "real-time chat app" → `realtime-chat`). Store it for later stages.22- `mkdir -p /job-queue/product-{name}/research-notes`23- Display: "🚀 Stage 0/6: Workspace created at /job-queue/product-{name}/"2425---2627## Stage 1: Big Idea & High-Level Research28291. **Analyze input.** From documentation: extract features, user flows, requirements, technical hints/constraints. From a description: core functionality, target users, key features, scale expectations.302. **Ask 3 initial clarifying questions** via `AskUserQuestion` — Product Vision, Scale & Audience, Deployment Context. Exact question/option text: `references/templates/stage1-big-idea.md`.313. **High-level technology research** via `WebSearch`: "[product type] architecture patterns 2026", "[product type] technology stack comparison 2026", "[product type] best practices 2026", "[scale level] architecture considerations 2026". `WebFetch` the top 2-3 articles per search; extract common patterns, recommended technologies, pitfalls. Save notes to `/job-queue/product-{name}/research-notes/stage1-research.md`.324. **Write `/job-queue/product-{name}/big-idea.md`** using the content structure in `references/templates/stage1-big-idea.md` (overview, value proposition, architecture approach, frontend/backend/database/deployment options with pros/cons, key challenges, next steps, research sources).335. Display: "✅ Stage 1/6: Big idea generated at product-{name}/big-idea.md"3435---3637## Stages 2-5: Parallel Deep Research for Architectural Documents3839Launch **all 4 research agents at once** via the `Task` tool with `subagent_type="general-purpose"`. Each agent researches one architectural document independently.4041| Stage | Task description | Output document | Prompt template |42|---|---|---|---|43| 2 | "Runtime execution research" — how the system executes work | `runtime-execution.md` | `references/templates/stage2-runtime-execution.md` |44| 3 | "Abstraction layer research" — how user intent becomes executable logic | `abstraction-layer.md` | `references/templates/stage3-abstraction-layer.md` |45| 4 | "Integration layer research" — how the system connects to external resources | `integration-layer.md` | `references/templates/stage4-integration-layer.md` |46| 5 | "Output rendering research" — how results are delivered to consumers | `output-rendering.md` | `references/templates/stage5-output-rendering.md` |4748**Build each agent prompt from its template:**49501. Read the stage's template plus `references/templates/common-closing-sections.md` — the single deduplicated tail (Technology Comparison / Implementation Considerations / Research Sources) shared by all four documents; insert it at the template's `[COMMON CLOSING SECTIONS]` marker with that stage's variants.512. Substitute `[Product Name]` and `{name}`, and paste the full big-idea.md content into the CONTEXT block (every agent needs it for cross-referencing).5253Each template already instructs the agent to WebSearch/WebFetch, ask its own `AskUserQuestion` clarifications, and write both its document and raw notes under `/job-queue/product-{name}/`.5455**Wait for all 4 agents to complete**, then display "✅ Stage N/6: [topic] research complete" for stages 2-5.5657---5859## Stage 6: Final Review & Presentation60611. Read all 5 generated documents (big-idea.md + the 4 architecture docs).622. Write `/job-queue/product-{name}/ARCHITECTURE-SUMMARY.md` (stack, style, key decisions, trade-offs, risks, next steps) and print the completion banner — both templates in `references/templates/stage6-summary-and-review.md`.633. Ask for feedback via `AskUserQuestion` (exact text in the same template): Approve, or Revise runtime execution / abstraction layer / integration layer / output rendering.64 - **Approve** → display "✅ Architecture approved! Ready for implementation."65 - **Revise** → re-run that stage's research agent with the feedback, regenerate the document, re-present for approval.6667---6869## Error Handling7071On any stage failure: display which stage failed, list completed vs failed stages, show the error, note partial work saved at `/job-queue/product-{name}/`, and give recovery instructions (resume with `/new-product [input]` or continue manually). Example failure display: `references/templates/stage6-summary-and-review.md`.7273---7475## Notes7677- **Tools**: WebSearch + WebFetch (deep research — use extensively), Task (parallel agents), AskUserQuestion (iterative clarification), Read/Write (documents).78- Ask clarifying questions during each stage, not all upfront; each agent reads big-idea.md for context. Documents must be comprehensive, well-structured, actionable.79- Duration ~20-40 min: Stage 0 seconds; Stage 1 ~5-10 min; Stages 2-5 ~15-25 min (parallel); Stage 6 ~2-3 min.80- Done = 6 documents in `/job-queue/product-{name}/` (big-idea, 4 architecture docs, ARCHITECTURE-SUMMARY), trade-offs and risks identified, user approval obtained.
artsmc