Home / onewave-ai / claude-skills · jev-eval/SKILL.md · GitHub

jev-eval skillA

jev-eval is agent-read markdown (skill) from onewave-ai/claude-skills: Build and run a labelled eval set for a System One model (Jev, Von, or any typed-decision config), then sweep criteria wordings and thresholds against it. Use when a Jev/Von classification is wrong or unreliable, when choosing between the hosted API and a local open model, when tuning noul thresholds, or before shipping any typed-decision feature. Produces an accuracy-by-wording matrix and a calibrated threshold..

Indexed from public GitHub and served as immutable, content-addressed versions. Install it pinned to an exact SHA-256 with the mdr CLI, and every file is verified against the hash recorded here before it reaches your agent. The deterministic audit below grades the latest version, and the same file always earns the same grade.

What the file says

# Evaluating a typed-decision config

**The eval set is the product.** A System One model's accuracy is dominated by how the
question was written, and the failure mode is silent — it returns a confident, type-valid,
wrong answer. Without labels you cannot tell a bad question from a bad model.

Measured: rewriting the criteria moved an open model from **4/15 to 14/15** on 15 records.
No model change. Then the same comparison at 150 records put that model at **61%** overall
against Jev's **97%** — the 15-record read was an artifact of a small, easy set. Both facts
are the point: wording swings results, and small sets lie about which way.

## Run it

```bash
python ~/.claude/skills/jev-eval/scripts/sweep.py labelled.json configs.json \
    --backend jev|von --question <name>
```

`labelled.json` is `[{"id","state","truth"}]`. `configs.json` maps a config name to
`{"instructions", "criteria"}` — a dict of options makes it a **choice**, a list of levels
makes it a **score**, omitting it makes it a **noul**. The script reads the Jev key from
Keychain (`typesafe-api-key`), prints accuracy per config, labels the spread
…

Read the whole file at its exact version.

How to install

Latest version
mdr add onewave-ai/claude-skills/jev-eval@git:20260923.352d573
Exact content
mdr add onewave-ai/claude-skills/jev-eval@sha256:c2f7fe3f473e83d1

Pin to a label to follow the author's releases, or to a sha256 to freeze the exact bytes forever. Either way the resolved hash is written to mdr.lock, and mdr install reproduces it on any machine.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_qkspczuhkor63vwg.svg)](https://markdownregistry.com/a/art_qkspczuhkor63vwg)

1 badge views in 30 days

Versions

versioncommittedcommitsizeaudit
git:20260923.352d573 latest2026-09-23 352d573 6,572 BA view · diff
git:20260921.1fa61752026-09-21 1fa6175 5,441 BA view

Audit of the latest version

A  17 of 17 checks passed. Deterministic, no model, same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (6572 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

onewave-ai/claude-skills · 306 stars · license MIT · pushed 2026-09-23 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_qkspczuhkor63vwg
GET https://markdownregistry.com/api/v1/resolve?ref=onewave-ai/claude-skills/jev-eval
GET https://markdownregistry.com/api/v1/blob/c2f7fe3f473e83d1c8274ce754a5c5ae6fedb3a076787c3537c3645b5f17143a

Your agent does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

More from onewave-ai/claude-skills

accessibility-auditor skill
onewave-ai/claude-skills · accessibility-auditor/SKILL.md · Audit websites for accessibility issues and WCAG compliance. Use when checking accessibility, fixing a11y issues, or…
git:20260608.d82d20f · audit A · 306 stars
agent-army skill
onewave-ai/claude-skills · agent-army/SKILL.md · Deploy a 2-layer parallel agent hierarchy for large, parallelizable work — big refactors, multi-file migrations…
git:20260704.01b429a · audit A · 306 stars
agent-eval-harness skill
onewave-ai/claude-skills · agent-eval-harness/SKILL.md · Use to test whether an AI agent, prompt, or skill actually works before or after it ships — build a small eval set…
git:20260920.97b5147 · audit B · 306 stars
agent-swarm-deployer skill
onewave-ai/claude-skills · agent-swarm-deployer/SKILL.md · Deploys swarms of sub-agents for massive parallel data processing tasks. Unlike agent-army (which is for code changes)…
git:20260608.d82d20f · audit A · 306 stars
agent-team-builder skill
onewave-ai/claude-skills · agent-team-builder/SKILL.md · Designs and deploys custom agent teams for specific business workflows. Interactive discovery of business processes…
git:20260608.d82d20f · audit A · 306 stars
agent-to-agent skill
onewave-ai/claude-skills · agent-to-agent/SKILL.md · Agent-to-Agent (A2A) communication protocol. Connect two or more Claude agents that pass messages, share context…
git:20260608.d82d20f · audit A · 306 stars
ai-coding-guardrails skill
onewave-ai/claude-skills · ai-coding-guardrails/SKILL.md · Use when an AI coding agent is writing or refactoring production code and its known failure modes need to be held in…
git:20260920.97b5147 · audit A · 306 stars
ai-readiness-assessment skill
onewave-ai/claude-skills · ai-readiness-assessment/SKILL.md · Assesses how ready a business is for AI adoption across six dimensions. Evaluates data maturity, tech stack, team…
git:20260608.d82d20f · audit A · 306 stars
animate skill
onewave-ai/claude-skills · animate/SKILL.md · Generate animated videos and motion graphics from natural language descriptions. Creates a standalone Vite + React…
git:20260225.059b9f3 · audit A · 306 stars
api-documentation-writer skill
onewave-ai/claude-skills · api-documentation-writer/SKILL.md · Generate comprehensive API documentation including endpoint descriptions, request/response examples, authentication…
git:20260608.d82d20f · audit A · 306 stars
api-endpoint-scaffolder skill
onewave-ai/claude-skills · api-endpoint-scaffolder/SKILL.md · Generate REST API endpoints with proper structure, validation, error handling, and types. Use when creating new API…
git:20260225.309ae7a · audit A · 306 stars
api-load-tester skill
onewave-ai/claude-skills · api-load-tester/SKILL.md · Load tests API endpoints with progressive concurrency. Measures response times, error rates, throughput, and identifies…
git:20260608.d82d20f · audit A · 306 stars

Every file in onewave-ai/claude-skills

Other files named jev-eval

jev-eval skill
softtor/nestjs-hexagonal · skills/jev-eval/SKILL.md · Calibrates or evaluates a semantic (Jev) rule of the nestjs-hexagonal rulebook — adds labelled golden cases, runs the…
git:20260921.98d05f3 · audit A · 5 stars

Browse by kind, by grade A, or by owner.