Software Architecture And Technical Planning Prompts

INPUTS
[requirements]; [scale]; [team]; [existing stack]; [budget]; [constraints]

PROMPT
Compare three architecture options appropriate to the requirements. Explain components, data flow, operational burden, failure modes, and tradeoffs in complexity, cost, and future change. Use realistic scope and avoid assuming extreme scale without evidence. Recommend one option with conditions and the strongest reason against it. Provide a compact component description and decision table. End with unknowns that materially affect the choice and a small prototype or measurement that could resolve the most important uncertainty.

INPUTS
[feature]; [requirements]; [system context]; [constraints]; [acceptance criteria]

PROMPT
Write a technical design document for this feature. Include the problem, scope, user flows, data model, interfaces, failure handling, operational considerations, and rollout approach. Explain alternatives and major decisions rather than listing technologies alone. Identify backward compatibility needs and unresolved questions. Provide an implementation sequence and verification plan tied to acceptance criteria. Keep examples compatible with the supplied system. End with review questions for the team and a clear distinction between required design elements and optional future improvements.

INPUTS
[use cases]; [entities]; [API style]; [consumers]; [authentication model]

PROMPT
Design an API contract for the supplied use cases. Define resources or operations, request and response schemas, validation, error structures, pagination, and idempotency where relevant. Clarify which fields are required and how missing or invalid values behave. Provide representative examples and versioning considerations. Do not assume authorization rules that are not specified. End with contract questions for consumers, compatibility risks, and checks that confirm implementations and clients agree on the most important boundary cases.

INPUTS
[schema]; [queries]; [business rules]; [database engine]; [scale assumptions]

PROMPT
Review the database schema against the business rules and expected queries. Evaluate entity boundaries, keys, relationships, constraints, nullable fields, and duplication. Identify concrete integrity risks and query patterns that may need indexing, without claiming performance improvements before measurement. Recommend focused changes with migration considerations. Provide a revised schema outline and validation queries. End with unresolved business definitions and a check that any proposed constraint is supported by actual requirements rather than assumptions about the data.

INPUTS
[current system]; [target system]; [data]; [downtime limits]; [dependencies]

PROMPT
Create a migration plan that preserves service continuity and data integrity. Define preparation, compatibility checks, data mapping, trial migration, reconciliation, cutover, monitoring, and rollback. Assign owners and decision gates to each stage. Identify irreversible steps and how to validate them before proceeding. Include a communication plan and a checklist for verifying critical workflows after migration. Use the supplied constraints and flag unknowns. End with go or no-go criteria and the evidence required to consider the migration complete.

INPUTS
[decision]; [context]; [options]; [constraints]; [chosen approach]

PROMPT
Write an architecture decision record that captures why the choice is reasonable now. State the context, requirements, considered options, decision, consequences, and assumptions. Explain benefits and drawbacks plainly, including what the decision makes harder. Link claims to supplied evidence and identify unknowns. Keep the document concise enough for future developers to understand without reading the discussion history. End with conditions that should trigger reconsideration and a review date or event. Do not rewrite tradeoffs as guaranteed benefits.

INPUTS
[system description]; [critical flows]; [incidents]; [availability goals]; [operations]

PROMPT
Assess reliability risks in the system's critical flows. Identify dependencies, single points of failure, timeouts, retries, overload behavior, and recovery processes where relevant. Distinguish observed weaknesses from hypotheses. Recommend practical safeguards in order of importance and explain their operational cost. Provide a failure-mode table and a verification plan. Avoid promising perfect availability. End with monitoring signals, escalation triggers, and the few recovery procedures the team should practice before relying on them during an incident.

INPUTS
[application]; [data]; [users]; [trust boundaries]; [deployment context]

PROMPT
Create a defensive threat model for the application. Identify assets, entry points, trust boundaries, plausible misuse, and consequences. Prioritize risks using the actual context rather than a generic checklist. Recommend prevention, detection, and recovery measures with owners and verification steps. Avoid exposing secrets or providing exploit instructions. Deliver a threat table and a concise summary of design changes. End with assumptions needing specialist review and a plan for checking that proposed controls address the identified risks without blocking legitimate use.

INPUTS
[product goals]; [technical debt]; [team capacity]; [dependencies]; [time horizon]

PROMPT
Create a technical roadmap that connects engineering work to product outcomes and operational needs. Group work into required capabilities, reliability improvements, and maintenance where relevant. Explain dependencies and sequence based on capacity rather than arbitrary dates. Define milestones, evidence of completion, and risks. Separate committed work from exploratory options. Provide a roadmap table and a review cadence. End with the decisions needed to make the roadmap realistic and the lower-priority items to defer if capacity is reduced.

INPUTS
[documentation inventory]; [system]; [user roles]; [support issues]; [known workflows]

PROMPT
Assess documentation gaps by the tasks readers need to complete. Identify missing, outdated, contradictory, or hard-to-find guidance. Compare needs of users, operators, and developers where applicable. Prioritize updates by frequency, failure impact, and maintenance effort. Provide a table of document, issue, audience, recommended change, and owner. Draft an improved outline for the highest-priority document. End with a review process that tests documentation through actual tasks rather than assuming a page is useful because it exists.