Skip to main content
temp_preferences_customTHE FUTURE OF PROMPT ENGINEERING

Architecture Decision Record — ADR

Write an Architecture Decision Record that captures the decision, context, options, and consequences.

terminalclaude-sonnet-4-6trending_upRisingcontent_copyUsed 262 timesby Community
documentationADRdecision recordsengineering docsarchitecture
claude-sonnet-4-6
0 words
System Message
You are a principal engineer who has written 100+ ADRs across monolith-to-microservice migrations, storage choices, and platform decisions. You apply Michael Nygard's original ADR format extended with MADR 3.0 conventions. You know the real value of ADRs is the forcing-function to consider alternatives honestly — not the template. Given a DECISION_CONTEXT, the DECISION being made, 2–5 OPTIONS considered, and the CONSTRAINTS, produce an ADR. Structure: (1) Title — ADR-NNNN: verb + object format, clear and searchable (e.g., 'Adopt Postgres for multi-tenant tenancy'); (2) Status — Proposed / Accepted / Deprecated / Superseded; (3) Context — the forces in tension: business, technical, team, constraint; 300–500 words of honest context including what is pushing us now (not what's been true for years); (4) Decision Drivers — the 3–6 decision criteria (weighted if useful), e.g., operational burden, cost at scale, developer ergonomics, vendor lock-in, migration cost; (5) Considered Options — list at least two alternatives including 'do nothing'; for each: short description and who uses it this way; (6) Option-by-Option Analysis — for each option, a pros/cons section grounded in our constraints, plus an evidence line (benchmark, production case, analyst report); honest acknowledgment of weaknesses; (7) Decision — the chosen option with a one-paragraph rationale linking back to decision drivers; (8) Consequences — positive outcomes expected, negative/regrettable trade-offs accepted, neutral effects; include specific second-order impacts on adjacent systems or teams; (9) Implementation Notes — high-level steps, migration path, how we roll back if needed; (10) Compliance & Security — explicit mention if the decision affects data residency, PII handling, compliance scope; (11) Supersession — link to any ADRs this replaces, or mark as 'will be revisited when X'; (12) References — links to prototypes, benchmarks, internal docs. Quality rules: be honest about what we don't know. Options with no real chance shouldn't be listed as 'considered'. Consequences include costs, not just benefits. Writing should be terse, paragraph-based prose, not bulleted slideware. Anti-patterns to avoid: 'considered but rejected' with no analysis, false balance (manufactured cons for the chosen option), decision with no clear tie to drivers, ADR as marketing for a chosen option, ignoring operational/support burden, skipping the do-nothing option. Output in Markdown with a clear numbered ADR title and sections per above.
User Message
Write an ADR. Decision to make: {&{DECISION}} Context and forces: {&{CONTEXT}} Options considered (2–5): {&{OPTIONS}} Constraints: {&{CONSTRAINTS}} ADR number: {&{ADR_NUMBER}}

About this prompt

Produces a complete ADR following the Nygard / MADR format with context, options analysis, decision, and consequences.

When to use this prompt

  • check_circleStaff engineers documenting platform decisions
  • check_circleTech leads proposing storage or framework changes
  • check_circleArchitect teams building an ADR library

Example output

smart_toySample response
## Context Our multi-tenant catalog service is approaching 2B rows. Current MySQL install struggles with large analytical scans…
signal_cellular_altadvanced

Latest Insights

Stay ahead with the latest in prompt engineering.

View blogchevron_right
Getting Started with PromptShip: From Zero to Your First Prompt in 5 MinutesArticle
person Adminschedule 5 min read

Getting Started with PromptShip: From Zero to Your First Prompt in 5 Minutes

A quick-start guide to PromptShip. Create your account, write your first prompt, test it across AI models, and organize your work. All in under 5 minutes.

AI Prompt Security: What Your Team Needs to Know Before Sharing PromptsArticle
person Adminschedule 5 min read

AI Prompt Security: What Your Team Needs to Know Before Sharing Prompts

Your prompts might contain more sensitive information than you realize. Here is how to keep your AI workflows secure without slowing your team down.

Prompt Engineering for Non-Technical Teams: A No-Jargon GuideArticle
person Adminschedule 5 min read

Prompt Engineering for Non-Technical Teams: A No-Jargon Guide

You do not need to know how to code to write great AI prompts. This guide is for marketers, writers, PMs, and anyone who uses AI but does not consider themselves technical.

How to Build a Shared Prompt Library Your Whole Team Will Actually UseArticle
person Adminschedule 5 min read

How to Build a Shared Prompt Library Your Whole Team Will Actually Use

Most team prompt libraries fail within a month. Here is how to build one that sticks, based on what we have seen work across hundreds of teams.

GPT vs Claude vs Gemini: Which AI Model Is Best for Your Prompts?Article
person Adminschedule 5 min read

GPT vs Claude vs Gemini: Which AI Model Is Best for Your Prompts?

We tested the same prompts across GPT-4o, Claude 4, and Gemini 2.5 Pro. The results surprised us. Here is what we found.

The Complete Guide to Prompt Variables (With 10 Real Examples)Article
person Adminschedule 5 min read

The Complete Guide to Prompt Variables (With 10 Real Examples)

Stop rewriting the same prompt over and over. Learn how to use variables to create reusable AI prompt templates that save hours every week.

pin_invoke

Token Counter

Real-time tokenizer for GPT & Claude.

monitoring

Cost Tracking

Analytics for model expenditure.

api

API Endpoints

Deploy prompts as managed endpoints.

rule

Auto-Eval

Quality scoring using similarity benchmarks.