jev-system-architect
Agent skill that finds fuzzy semantic judgments in a codebase and redesigns them as small Jev Choice, Score or Noul primitives.
This repository is mostly documentation or write-up rather than runnable code.
Overview
Jev System Architect is an architecture skill for Cursor, Claude Code and compatible agent harnesses. It complements TypeSafe's official skill, which supplies API and SDK details, by repeatedly asking where a system contains fuzzy semantic judgment that should become a small Jev decision primitive.
It applies to systems with unstructured data, brittle parsing, prompt-to-JSON workflows, semantic routing, classification, ranking, verification, guardrails or extraction from known candidates. The skill guides an agent through whether Jev belongs at all, where the boundary sits, which judgments become Choice, Score or Noul questions, what state each sees, which questions run together, how code composes results, how uncertainty changes behaviour and how to evaluate before production. Its mental model keeps control flow, permissions, calculations, side effects and policy in deterministic code. It explicitly tells agents not to invent API fields.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Other | Where in this system should fuzzy judgment become a Jev Choice, Score or Noul question? | — |
How it works
- Agent reviews the system for brittle semantic logic
- Identifies judgment points and the Jev boundary
- Designs bounded questions and the state each sees
- Plans code composition, uncertainty routing and evaluation
Strengths
- Focuses on architecture rather than API syntax
- Keeps side effects and policy in deterministic code
- Pairs with the official TypeSafe skill
Caveats
- Guidance only; no code or tests
- Quality depends on the host agent following it
Getting started
cp -R skills/jev-system-architect ~/.cursor/skills/jev-system-architect (or your agent's skills folder).At a glance
- Type
- Agent skill
- Status
- beta
- Runtime
- Agent skill
- Has code
- no
- Has tests
- no
- License
- MIT
Use cases and integrations
How curated lists describe it
Skill that hunts for brittle semantic logic and turns it into Choice / Score / Noul boundaries.
AnotiaWang/awesome-decision-modelsSystem-architecture skill for TypeSafe AI Jev/System One — find fuzzy semantic…
wh000wh000/awesome-jev-liveSystem-architecture skill for TypeSafe AI Jev/System One — find fuzzy semantic judgment and turn it into small Choice/Score/Noul primitives.
hellogumbo/awesome-jev
Frequently asked questions
How is this different from TypeSafe's official skill?
The official skill covers API and SDK usage; this one focuses on where Jev belongs in a system and how to decompose judgments.
Does it write Jev calls for me?
It guides the agent's design; for exact syntax it defers to TypeSafe's docs and official skill.
Which agents can use it?
Cursor, Claude Code and other harnesses that load SKILL.md files.
Related projects
- Building with Jev★ 155
Agent skill for writing and improving programs that call Jev: question design, state structure, thresholds and diagnosing wrong answers.
Agent skillproduction Portable Agent Skill teaching coding agents to use Jev as a bounded judgment layer, plus an optional MCP bridge.
MCP serverbetaTypeScriptIndependent agent skill that makes a coding agent design, build, test and tune a bounded Jev decision layer with a contract first.
Agent skillbetaAgent skill that guides you in finding, designing and testing Jev decision loops, including when to keep logic in code.
Agent skillbetaClaude Code and Codex skills plus a study course for building with Jev, including Jevaluate, which rates how well projects use Jev.
ProjectbetaCommunity Jev skill for Claude Code and Codex: a guide, a dependency-free Python CLI and 5 measured question sets.
Agent skillbetaPython