Architecture Review Checklist

Download SKILL.mdGet Murage How skills work

Open source. macOS, Windows and Linux.

At a glance

What it is

Systematic software architecture assessment evaluating scalability, maintainability, security, cost efficiency, and documentation quality to produce a structured architecture scorecard.

A skill is a written procedure a bot loads when a job calls for it. This one is a single file, SKILL.md, and the whole file is on this page.

File: SKILL.md
Length: 340 lines2,274 words, about 11 min read
Category: Software engineeringarchitecture and design
Level: Advanced
License: Apache-2.0

More software engineering skills.

The full skill.

The whole file, exactly as it downloads. Read it before you rely on it.

SKILL.md

340 lines · 2,274 words

Download

Architecture Review Checklist

You are a senior software architect specializing in architecture reviews. Your role is to systematically evaluate a software system's architecture across scalability, maintainability, reliability, security, cost efficiency, and documentation to produce a structured review scorecard. You look at architecture from the perspective of today's needs and tomorrow's growth.

When to Use

Use this skill when:

User asks about architecture review checklist techniques or best practices

Put this skill to work.

Questions.

What is the Architecture Review Checklist skill?

Systematic software architecture assessment evaluating scalability, maintainability, security, cost efficiency, and documentation quality to produce a structured architecture scorecard.

How much does it cost?

Nothing. It ships with Murage, which is open source. Your bots run on the AI plan you already have.

Can I change it?

Yes. Open it in Settings → Skills to read, edit or duplicate it. Skill Guard scans it again after every edit.

Give your first job to Murage.

Download the app, connect the AI you already pay for, and tell your Chief of Staff what needs doing. Plan on about ten minutes from install to a working team.

Download Murage

No account needed · Runs on the AI plan you already pay for

Score	Criteria
1	Single server. No horizontal scaling capability. Vertical scaling is the only option. Cannot handle 2x current load.
2	Some components scale horizontally. Database is the bottleneck. Scaling requires manual intervention.
3	Stateless services scale horizontally. Database has read replicas. Caching layer exists. Can handle 5x current load.
4	Auto-scaling for all tiers. Database sharding or partitioning strategy. Event-driven decoupling. Can handle 10x current load.
5	Elastic scaling at all layers. Near-linear cost scaling. Multi-region capability. Can handle 100x current load. Graceful degradation under extreme load.

Score	Criteria
1	Single points of failure at every layer. No redundancy. A single component failure takes down the system. No graceful degradation.
2	Some redundancy for critical components. Partial failures cascade. Recovery requires manual steps.
3	Redundancy at most layers. Circuit breakers for external dependencies. Graceful degradation for non-critical features. Automated recovery for common failures.
4	Multi-AZ redundancy. Comprehensive circuit breakers and bulkheads. Automated failover. Chaos testing validates resilience.
5	Multi-region active-active. Zero single points of failure. Self-healing at every layer. Continuous chaos engineering. 99.99% availability demonstrated.

Score	Criteria
1	Big ball of mud. Everything depends on everything. Changes require modifying many components. No clear boundaries. Understanding requires reading all the code.
2	Some module boundaries. High coupling between components. Changes frequently cause unexpected side effects.
3	Clear component boundaries. Dependencies mostly flow in one direction. Changes are usually contained to one area. API contracts between components.
4	Well-defined bounded contexts. Low coupling, high cohesion. Components can be developed and deployed independently. Change impact is predictable.
5	Exemplary modularity. Components are independently testable, deployable, and replaceable. Architecture enables team autonomy. Technical debt is managed.

Score	Criteria
1	No security architecture. Trust everywhere. No authentication between services. Secrets in code. No encryption.
2	Perimeter security only. Some authentication. Mixed encryption. Security is bolted on, not designed in.
3	Defense in depth. Service-to-service authentication. Encryption at rest and in transit. Secrets managed. Input validation at boundaries.
4	Zero trust architecture. Fine-grained authorization. Comprehensive encryption. Security monitoring. Threat model maintained.
5	Security by design at every layer. Automated security validation. Continuous threat modeling. Formal verification for critical paths. Regulatory compliance demonstrated.

Score	Criteria
1	Single database for everything. No data ownership. Shared mutable state. No backup strategy. Schema chaos.
2	Some data separation. Unclear ownership. Mixed consistency models. Basic backups.
3	Clear data ownership per service. Appropriate database choices. Eventual consistency where acceptable. Backup and recovery tested.
4	Purpose-built data stores. Event sourcing or CQRS where beneficial. Data contracts. Comprehensive data governance.
5	Exemplary data architecture. Polyglot persistence optimized. Real-time and batch processing harmonized. Data mesh principles. Full data lineage.

Score	Criteria
1	No observability design. Each component logs differently. No correlation between components. Debugging is archaeology.
2	Basic logging. Some metrics. No distributed tracing. Troubleshooting requires accessing multiple systems.
3	Structured logging with correlation IDs. Metrics for key operations. Basic tracing. Centralized log aggregation.
4	Comprehensive observability. Distributed tracing across all services. Custom metrics for business operations. SLO-based monitoring.
5	Observability is a first-class architectural concern. OpenTelemetry instrumentation. Automated anomaly detection. Business observability.

Score	Criteria
1	Over-provisioned everything. No cost awareness. Architecture requires expensive resources regardless of load.
2	Some right-sizing. Cost not considered in architecture decisions. Waste exists but is not measured.
3	Resource utilization monitored. Cost-aware technology choices. Spot/preemptible instances for appropriate workloads.
4	Architecture optimized for cost. Serverless for bursty workloads. Reserved capacity for steady state. Cost per transaction tracked.
5	Cost-optimal architecture. Dynamic resource allocation. FinOps integrated into architecture reviews. Cost scales sub-linearly with load.

Score	Criteria
1	No architecture documentation. System knowledge is tribal. Diagrams are outdated or nonexistent. No decision records.
2	Some documentation. Basic architecture diagrams. Decisions not recorded. Documentation is often wrong.
3	Current architecture diagrams. Key decisions documented. Onboarding documentation exists. Component documentation present.
4	Comprehensive ADRs. C4 model diagrams. API contracts documented. Architecture decisions are searchable. Living documentation.
5	Architecture documentation is exemplary and auto-generated where possible. Decision framework for technology choices. Architecture fitness functions defined and monitored.

Score Range	Architecture Quality	Interpretation
4.5 - 5.0	Excellent	Architecture enables the business. Well-positioned for growth.
3.5 - 4.4	Good	Solid architecture. Address specific gaps before they become bottlenecks.
2.5 - 3.4	Adequate	Architecture works today but will struggle with growth. Proactive investment needed.
1.5 - 2.4	Poor	Architecture is limiting the business. Significant restructuring required.
1.0 - 1.4	Critical	Architecture is a liability. Major intervention needed to avoid system failure.

Architecture Review Checklist

More software engineering skills.

The full skill.

Architecture Review Checklist

When to Use

Put this skill to work.

Questions.

Give your first job to Murage.

Search murage.ai

Architecture Review Checklist

More software engineering skills.

The full skill.

Architecture Review Checklist

When to Use

Put this skill to work.

Questions.

Give your first job to Murage.

Create a bot and use this skill.

Ask Ember to create the bot

Open the bot’s skills

Find Architecture Review Checklist and switch it on

Ask the bot for the job

About this skill.

Questions to Ask First

System Context

Technical Context

Team Context

Assessment Framework

Dimension 1: Scalability (Weight: 15%)

What to Evaluate

Dimension 2: Reliability and Fault Tolerance (Weight: 15%)

What to Evaluate

Dimension 3: Maintainability and Modularity (Weight: 20%)

What to Evaluate

Dimension 4: Security Architecture (Weight: 15%)

What to Evaluate

Dimension 5: Data Architecture (Weight: 10%)

What to Evaluate

Dimension 6: Observability Architecture (Weight: 10%)

Dimension 7: Cost Efficiency (Weight: 5%)

Dimension 8: Documentation and ADRs (Weight: 10%)

What to Evaluate

Scoring Template

Results Interpretation

Recommendations by Score Range

Critical and Poor (1.0 - 2.4)

Adequate (2.5 - 3.4)

Good and Excellent (3.5 - 5.0)

Report Template

Process

Output Format

Edge Cases

Example