Generate reviewable system technical design documents
Form a complete design from business goals to architecture, data, interfaces, reliability, security and migration solutions.
Traceable sourceFabric · create_design_document ↗Complete prompt
You are a pragmatic software architect. Please write a technical design document based on the following information. System or function: [Requirement description] Existing architecture: [Technology stack and current situation] Scale expected: [Number of users, requests or data] Key constraints: [time, cost, security, compliance or compatibility] Outputs: Background and goals, non-goals, architecture overview, component responsibilities, key processes, data models, interface contracts, conformance strategies, error handling, observability, security and privacy, capacity estimates, deployment and rollback, migration plan, testing strategy, trade-offs and alternatives, open issues. Prioritize reuse of existing capabilities. Explain the reasons for the choice, costs, and failure modes for each important design decision.
Example output
Solution overview: The database saves scheduling facts, the Redis delay queue is responsible for triggering, and the sending service uses idempotent keys to deduplicate. Key trade-off: Redis is not used as the sole source of truth to avoid permanent loss of tasks due to restarts.
USAGE GUIDE
When to use it
Form a complete design from business goals to architecture, data, interfaces, reliability, security and migration solutions. Use it as a structured starting point, then review the result against your original material.
How to fill the variables
- Requirement description
- Add concrete facts, context, and constraints. Mark unknown details instead of guessing.
- Technology stack and current situation
- Add concrete facts, context, and constraints. Mark unknown details instead of guessing.
- Number of users, requests or data
- Add concrete facts, context, and constraints. Mark unknown details instead of guessing.
- time, cost, security, compliance or compatibility
- Add concrete facts, context, and constraints. Mark unknown details instead of guessing.
Get better results
- Describe the audience and intended decision.
- Include source material and hard constraints.
- State what the model must not invent.
Before you use the output
- Verify facts, numbers, and quotations.
- Check that uncertainty is clearly labeled.
- Test or review high-impact recommendations.
Editorial note: This template has been structured for practical use. Results vary by model and input; verify important outputs against primary sources.