Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add OneWave-AI/claude-skills --skill jev-evalgit clone --depth 1 https://github.com/OneWave-AI/claude-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/onewave-ai/claude-skills/jev-eval)<a href="https://agentmods.dev/skills/onewave-ai/claude-skills/jev-eval"><img src="https://agentmods.dev/badge/skills/onewave-ai/claude-skills/jev-eval/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/onewave-ai/claude-skills/jev-eval"><img src="https://agentmods.dev/badge/skills/onewave-ai/claude-skills/jev-eval.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00092 | $0.01785 |
| Opus 5.5 | $0.00037 | $0.00714 |
| Sonnet 5.5 | $0.00018 | $0.00357 |
| Haiku 4.5 | $0.00009 | $0.00178 |
Grade A, and why
jev-eval scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 23d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 142 lines — stays where its author put it; the contents beside it link to each section on GitHub.
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
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
ROBUST or FRAGILE, sweeps thresholds for nouls, and scores the confidence gate.
Real output, 50 records of agent shell-command risk, only the backend changed:
config accuracy ms/rec
A terse one-liners 45/50 410 <- Jev
B rich criteria 49/50 415
C rich + exclusions 46/50 418
D deliberately lazy 44/50 411
spread: 44/50 to 49/50 (ROBUST - wording is not load-bearing)
A terse one-liners 9/50 66 <- Von, same configs
B rich criteria 23/50 121
C rich + exclusions 15/50 129
D deliberately lazy 22/50 53
spread: 9/50 to 23/50 (FRAGILE - and the ceiling is still not usable)
Read the ceiling before the spread. A FRAGILE model whose best config is 23/50 is not a wording problem you can write your way out of — it is the wrong model for that question.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 23d ago Changed · +16 lines c2f7fe3f473e
- 25d ago First seen · 126 lines · 92 tokens per session scan A 2e41d2360b6b
jev-eval is a skill published in the GitHub repository OneWave-AI/claude-skills (349 stars, last pushed 15d ago), licensed MIT. It adds 92 tokens to every session and 1,785 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-22.
Other skills, from other repositories
seo-aeo-geo
Optimize content and site structure for AI-driven search experiences including AI overviews, large language model citations, generative answer engines, and AI assistants. Use this skill whenever the user wants to optimize for AI search, get cited by language models, appear in AI overviews, build llms.txt, structure…
automl-hyperparameter-optimization
Best practices for AutoML and hyperparameter search with Optuna, Ray Tune, and PyCaret, covering search-space design, validation splits, and leakage prevention. Use when tuning model hyperparameters, setting up a pruned or distributed hyperparameter search, designing a nested validation scheme, or evaluating whether…
google-adk
Best practices for building AI agents with Google's Agent Development Kit (ADK) in Python, covering agent design, tools, sessions, memory, artifacts, evaluation, and deployment. Use when building LLM agents or multi-agent systems with ADK, defining ADK tools, wiring up sessions/state/memory, working with ADK…
pyspark-etl
Best practices for building performant, testable PySpark ETL pipelines with Spark SQL and Apache Iceberg. Use when writing or reviewing PySpark jobs, designing joins and window functions, working with map/array higher-order functions, or building idempotent cumulative/snapshot table merges.
anthropic-claude-development
Expert guidance for Anthropic Claude API development including Messages API, tool use, prompt engineering, and building production applications with Claude models.
openai-api-development
Expert guidance for OpenAI API development including GPT models, Assistants API, function calling, embeddings, and best practices for production applications.