well-architected-review
Skill1 starsChanged 3 months agoMITView on GitHub
Run a Well-Architected review of a workload across reliability, security, cost optimization, operational excellence, performance efficiency and sustainability. Use when the user asks for a WAF review, architecture assessment, production-readiness check, design review before go-live, or wants to know if a workload is fit for production.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- ricmmartins/azure-sre-agent-skillstheir own version, first seen Nov 15, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- ricmmartins/azure-sre-agent-skills★ 70same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository4 other files it gives its agents.
- finops-reviewSkill
- landing-zone-reviewSkill
- terraform-security-reviewSkill
- threat-modelSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 97
If the workload is not production-ready, say so in the first line.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Well-Architected Review
Assess one workload, not a whole estate. Pillars trade against each other — name the trade-offs the design has already made, and whether they were deliberate.
Method
- Establish the workload's criticality tier, RPO/RTO, and expected load. Without these, most pillar questions have no correct answer — ask first.
- Assess each pillar against the questions below.
- Rate: Strong / Adequate / Weak / Not assessed.
- Finish with the trade-off register: what this design sacrifices, and whether the sacrifice matches the business requirement.
Reliability
- Single points of failure across compute, data, network, identity and DNS
- Availability zone and region strategy; is the failover tested or theoretical?
- RPO/RTO defined per tier and evidenced by an actual restore test
- Health modelling: what "healthy" means, and how the system knows
- Retry, timeout, circuit-breaker and back-off on every external dependency
- Graceful degradation: what still works when a dependency is down
- Capacity headroom and autoscaling limits; what happens at the ceiling
- Dependency on any single team, key, or subscription/account quota
Security
- Identity as the primary perimeter; workload identities not static credentials
- Least privilege actually enforced, not aspirational
- Network segmentation and private connectivity to data services
- Encryption at rest and in transit; key ownership and rotation
- Secrets lifecycle: issuance, rotation, revocation
- Input validation and injection defence at every trust boundary
- Detection: are the logs that would reveal a breach actually collected?
- Patch and vulnerability management for the whole stack, including base images
Cost optimization
- Right-sizing evidence, not guesswork; utilisation data reviewed
- Commitment coverage (reserved/savings plans) matched to steady-state baseline
- Autoscaling scales down, and non-production shuts off out of hours
- Storage tiering and lifecycle policies for cold data
- Egress cost paths identified — usually the surprise line item
- Cost attribution by tag; a named owner per component
- Cost of resilience stated explicitly (what multi-region actually costs)
Operational excellence
- Everything deployed from code; no manual production changes
- Deployment strategy: blue/green, canary or rolling — with a tested rollback
- Observability covers the three signals plus business KPIs
- Alerts are actionable and tied to user-visible symptoms, not raw resource metrics
- Runbooks exist for the top failure modes and have been used
- On-call: rotation, escalation, and a blameless postmortem practice
- Change failure rate and lead time known, not guessed
Performance efficiency
- Performance targets defined (latency percentiles, throughput), not "fast"
- Load tested at and beyond expected peak; the breaking point is known
- Data access patterns match the store chosen; indexes and partitions verified
- Caching strategy with an explicit invalidation rule
- Async and queue-based decoupling where latency budgets allow
- Cold-start and connection-pool behaviour under burst
Sustainability
- Workload placement and scheduling relative to demand
- Resource utilisation targets; idle capacity treated as waste
- Data retention proportionate to actual need
- Managed services preferred over self-hosted where they consolidate utilisation
Output
## Well-Architected Review — <workload>
Criticality: <tier> RPO: <x> RTO: <y>
| Pillar | Rating | Headline |
|---|---|---|
| Reliability | | |
| Security | | |
| Cost optimization | | |
| Operational excellence | | |
| Performance efficiency | | |
| Sustainability | | |
### Top risks before production
1. <risk> — likelihood, impact, mitigation, effort
### Trade-off register
| Trade-off | Chosen | Deliberate? |
### Not assessedIf the workload is not production-ready, say so in the first line.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

