doanchienthangdev

Muse:Benchmark

Name: Muse:Benchmark
Author: doanchienthangdev

Development community

SOURCE https://github.com/doanchienthangdev/muse/blob/main/commands/muse:benchmark.md

Description

--- description: Muse persona benchmark — measures distinctiveness, voice discipline, schema compliance, and mode fit across all 8 personas. Runs static checks (C1-C12, Jaccard overlap, schema mirror) + subagent-based blind Turing simulation on sample prompts from benchmarks/universal.md. Writes markdown report + JSON to ~/.muse/benchmark-reports/. Supports --baseline, --diff, --quick, --persona, --trend modes. Regression detection against baseline.json. allowed-tools: Read, Glob, Bash, Write, E

Installation

Installs to ~/.claude/skills/doanchienthangdev-muse-muse-benchmark/SKILL.md

Terminal

mkdir -p ~/.claude/skills/doanchienthangdev-muse-muse-benchmark && curl -fsSL https://raw.githubusercontent.com/doanchienthangdev/muse/HEAD/commands/muse:benchmark.md -o ~/.claude/skills/doanchienthangdev-muse-muse-benchmark/SKILL.md

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