How do I know if the skill is stale against the live API?

Run scripts/verify.sh with a key; it replays examples/verified/ and fails on schema drift instead of silently passing.

Can I use the Java example without a framework?

Yes. The README describes it as a dependency-free client in examples/JevClient.java, separate from the skill prose.

Does installing it change how Claude Code behaves?

Claude Code picks it up automatically after cloning to ~/.claude/skills/jev-expert, callable as /jev-expert or triggered by TypeSafe/Jev work.

Will this teach me when not to use Jev?

That is explicit scope: it covers where Jev is weak and confidence quirks, not just happy-path classification and routing.

How do I install the jev-expert skill?

Clone the repository into ~/.claude/skills/jev-expert. Claude Code picks it up automatically, and you can call it with /jev-expert.

What topics does the jev-expert skill cover?

The API contract, the Choice, Score and Noul primitives, confidence calibration and its quirks, where Jev is weak, integration patterns such as classify, route, rerank, gate and guarding an LLM, real-time loops, and a dependency-free Java client at examples/JevClient.java.

Does the skill trigger without a slash command?

Yes. Besides being callable as /jev-expert, it also triggers on any TypeSafe or Jev work in Claude Code, so it can load on its own when a task touches the API.

CommunityJev Agents & MCP

Jev Expert

Claude Code skill documenting Jev typed decisions, calibration quirks, and classify-route-rerank-gate patterns, with a dependency-free Java client and replayable fixtures.

See highlights

Open github.com

Stars: 0
Language: Shell
Published by: Community
Author: slashdaemon
Last verified: Oct 2026

slashdaemon/jev-expert

User repository on GitHub

View on GitHub

A Claude Code skill: an authoritative reference for TypeSafe AI's Jev, the "System One" model. You send it a state plus typed questions (Choice / Score / Noul) and it returns typed answers with calibrated probabilities. It doesn't generate text. The skill covers the API contract, primitives, confidence calibration and its quirks, where Jev is weak

Forks: 0
Last pushed: Updated Sep 2026

Repo stats from the GitHub API, cached Oct 2026.

About this project

A Claude Code skill: an authoritative reference for TypeSafe AI's Jev, the "System One" model. You send it a state plus typed questions (Choice / Score / Noul) and it returns typed answers with calibrated probabilities. It doesn't generate text. The skill covers the API contract, primitives, confidence calibration and its quirks, where Jev is weak, integration patterns (classify, route, rerank, gate, guard an LLM), real-time loops, and a dependency-free Java client (examples/JevClient.java).

Quoted from the project's README. Read the full README.

Tagscommunity github

Works forJev for Scoring

Used inJev for Software & DevTools

Languages: Shell

Highlights

Covers API contract, primitives, confidence calibration quirks, and where Jev is weak.
Documents classify, route, rerank, gate, guard-an-LLM, and real-time loop integration patterns.
Ships dependency-free examples/JevClient.java plus verified fixtures checked by scripts/verify.sh.

Watch out

Skill text alone needs no key, but replaying fixtures needs TYPESAFE_API_KEY and fails on schema drift. Copy reviewed against the repository README on 2026-10-04, not from running it.

Questions about this entry

Short answers drawn from the linked project or page.

More Jev resources like this

jev-mcp

MCP server that exposes Jev classify, score, yes-no check, risk gate, and multi-question decide tools to coding agents through one Jev call per tool.

Jev Agents & MCP#community#github

Community

codaaiteam/jev-mcpJavaScript

jevless

Independent toolkit that reproduces Jev-style Choice, Noul, and Score decisions on any model exposing log-probabilities, plus a local server speaking the same systemone request shape.

Jev Agents & MCP#community#github

Community

jbpayton/jevlessPython

jev-decision-maker-at-meteora

Condor drop-in where Jev Noul fan-out ranks Meteora pools and a Score sizes positions, while math clamps enforce rug cards, cost tiers, dry-run, and one-sided walls.

Jev Agents & MCP#community#github

Community

cedarmuse-creator/jev-decision-maker-at-meteoraPython

jev-agent-harness

Beginner-friendly harness guide showing Jev as the fast classifier inside agent loops, with ModelRouter and AutoMode middleware examples.

Jev Agents & MCP#community#github

Community

bpmforbusiness/jev-agent-harness

jev-code

Five typed tools — classify, check, score, rank, and ask — plus a skill, wired into Claude Code, Codex, Pi, and OpenCode in one setup command.

Jev Agents & MCP#community#typescript#mcp

Community

FrancoisChastel/jev-codeTypeScript

jev-reranking

Reproducible study testing whether Jev zero-shot can match specialist rerankers on fixed MS MARCO passage, MS MARCO document, and TREC-1 WSJ candidates, reporting ranking quality, time, and API cost.

Jev Agents & MCP#community#github

Community

carlaiau/jev-rerankingPython

Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai