Adversarial reviewer personality for architecture discussions. Use when a user requests a design review, architecture review, system design critique, tech stack decision, RFC review, or devil's advocate perspective on trade-offs. Makes Claude challenge assumptions instead of agreeing — questioning scalability assumptions, identifying single points of failure, challenging technology choices, and probing for edge cases rather than validating decisions.
97
100%
Does it follow best practices?
Impact
94%
1.25xAverage score across 5 eval scenarios
Passed
No known issues
A 3-person engineering team at a media company is building an internal content scheduling tool. The tool allows editors to plan, draft, and schedule social media posts across a handful of company accounts. Expected users: approximately 15 internal editors. No public-facing traffic.
The team has produced an architecture proposal and wants a design review before they start building.
Write your review to review.md.
The following files are provided as inputs. Extract them before beginning.
=============== FILE: proposal.md ===============
An internal tool for scheduling social media content. 15 editor users. Posts are queued and published automatically at scheduled times.
We will build a microservices system with the following services:
Each service will be independently deployed on Kubernetes (GKE). We will use Istio as the service mesh to handle inter-service communication, mTLS, and traffic management.
We will instrument all services with distributed tracing using Jaeger to make debugging easier across the service graph.
This approach gives us flexibility to scale individual services independently as usage grows, and follows modern microservices best practices.