jstuart0

Mozart Eval — Data skill for Claude Code

Data community

Run mozart's EVAL pipeline — evaluate mozart's own field performance from past campaign artifacts across your projects, verify whether prior fixes actually worked, and propose configuration improvemen.

How to install Mozart Eval

Installs to ~/.claude/skills/jstuart0-mozart-orchestration-mozart-eval/SKILL.md

Terminal

mkdir -p ~/.claude/skills/jstuart0-mozart-orchestration-mozart-eval && curl -fsSL https://raw.githubusercontent.com/jstuart0/mozart-orchestration/HEAD/commands/mozart-eval.md -o ~/.claude/skills/jstuart0-mozart-orchestration-mozart-eval/SKILL.md

Restart Claude Code, or start a new session, for it to be picked up.

What Mozart Eval does

description: Run mozart's EVAL pipeline — evaluate mozart's own field performance from past campaign artifacts across your projects, verify whether prior fixes actually worked, and propose configuration improvements. Delta-scoped via a persistent eval ledger so repeat runs only examine what changed.

/mozart-eval — evaluate and improve mozart from field evidence

You are now mozart in EVAL shape for this session. The subject is mozart itself: the campaign artifacts (state files, flow s

Alternatives in Data

Eval Skills — Eval all skills with sufficient data, rank by procedure-following score, identify candidates for optimization 31 ★
Crune — Decipher the traces etched in past sessions and resurrect them as reusable skills 23 ★
Marketing Roas — CorpusIQ prompts for marketing performance — CAC, ROAS by channel, campaign analytics across Google Ads, Meta 18 ★