Repeatable client analysis
Run the same structured analysis for every client or business unit so quality no longer depends on who picked up the work.
— Product & Tech
Reads architecture documentation, diagrams, design decisions and any incident or performance material, then reviews the design against four lenses: scalability, resilience and availability, security by design, and accumulated technical debt.
2–3 weeks → ~25 minutes
For a typical multi-document review workflow
No coding required
— USE CASES
Run the same structured analysis for every client or business unit so quality no longer depends on who picked up the work.
Turn raw source material into a working draft in minutes and spend your time on judgement instead of assembly.
Produce the same shape of output each cycle so movement is measurable rather than re-argued.
— HOW IT BEHAVES
The mechanics behind this specific template — what it reads, what it calculates, and where a human stays in the loop.
Each decision is assessed against criteria you control and weight, so the same standard applies to every decision in the decision set in scope.
Every decision is captured in the same field structure, so records drawn from different documents and sources stay comparable.
All of your governance and authority documentation is processed as one set, so patterns across documents surface instead of being read one file at a time.
Findings on the decision set in scope are written up as a document that reads like professional output, with each claim tied back to a decision.
Every product & tech team needs technical architecture review — and almost none of them do it the same way twice. In practice it shows up as repeatable client analysis: run the same structured analysis for every client or business unit so quality no longer depends on who picked up the work. The value sits in the rigour, not the typing — yet the rigour is exactly what gets traded away when there is only 2–3 weeks of capacity for it.
Skillsize turns that work into a Skill: you supply the material, and what comes back is a structured, review-ready deliverable, with a traceable record of what informed each conclusion. The judgement is built in — how items are broken up, what standard they are held to, and where the run stops for a human review. In effect, 2–3 weeks of senior time compresses into ~25 minutes — and the output is comparable across clients, quarters and colleagues instead of shaped by whoever ran it.
Technical architecture review exports as a structured SKILL.md file and is MCP-ready, so the same method runs in ChatGPT, Claude, Copilot or your own AI products. Adapt it to your methodology, and the intelligence stays yours — not locked to one vendor.
Data, AI & technology governance review
Reads data-sharing and customer agreements, AI/model inventory and documentation, governance policies and any DPIA / model risk assessments.
Security & compliance posture scan
Reads pen-test and security review reports, SOC 2 / ISO evidence, vendor security questionnaire responses and the incident log.
Tech stack & architecture inventory
Reads the target's technology documentation, architecture diagrams and vendor list, extracts every system with its vendor, purpose, integrations, hosting and build-or-buy status, then researches each vendor's viability, ownership changes and end-of-life risk.