experience-score-calibrator — Claude Code skill: score calibration
experience-score-calibrator is a Claude Code skill that calibrates Experience Score dimensions against reviewed evidence. It checks coverage and uncertainty before presenting an aggregate score.
MIT ★ 529 updated 2026-11-17
npx skills add ShuaiMXu/happyhands --skill experience-score-calibrator -a claude-codeWhat experience-score-calibrator does
The skill compares dimension results with human judgments, historical cases, evidence coverage, and uncertainty. It calibrates dimensions before considering the total, then reports confidence and missing evidence.
Its output contains structured evidence, findings, proposed updates, and limitations. When required inputs or professional review are missing, it returns needs-evidence. It does not invent an Experience Score or turn one project observation into a general rule. A score jump must not be published without a validated rubric and sufficient cases.
What it helps with
- compare Experience Score dimensions with reviewed evidence
- show confidence and gaps in the evidence
- draft rubric updates with findings and limitations
- withhold an unsupported score jump
Example prompts for the agent
- Calibrate the Experience Score dimensions against the results, human judgments, and historical cases.
- Show the confidence level and missing evidence for this score.
- Prepare structured findings and proposed rubric updates; flag any missing evidence.
How to install experience-score-calibrator
1Skills CLI (Vercel)
npx skills add ShuaiMXu/happyhands --skill experience-score-calibrator -a claude-code2Manually: put the files in place
~/.claude/skills/experience-score-calibrator/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/experience-score-calibrator/
1Skills CLI (Vercel)
npx skills add ShuaiMXu/happyhands --skill experience-score-calibrator -a codex2Manually: put the files in place
.agents/skills/experience-score-calibrator/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add ShuaiMXu/happyhands --skill experience-score-calibrator -a cursor2Manually: put the files in place
.cursor/skills/experience-score-calibrator/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install ShuaiMXu/happyhands skills/product-design/experience-score-calibrator2Manually: put the files in place
.github/skills/experience-score-calibrator/SKILL.md1Skills CLI (Vercel)
npx skills add ShuaiMXu/happyhands --skill experience-score-calibrator -a gemini-cli2Platform CLI
gemini skills install https://github.com/ShuaiMXu/happyhands --path skills/product-design/experience-score-calibratorManually: put the files in place
.devin/skills/experience-score-calibrator/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about experience-score-calibrator
How do I install experience-score-calibrator in Claude Code?
Run npx skills add ShuaiMXu/happyhands --skill experience-score-calibrator -a claude-code. The skill file is at ~/.claude/skills/experience-score-calibrator/SKILL.md.
What does experience-score-calibrator do?
It calibrates Experience Score dimensions against reviewed evidence and reports confidence and gaps before presenting an aggregate score. Its output includes findings, proposed updates, and limitations.
What inputs does experience-score-calibrator need?
It needs dimension results, human judgments, evidence coverage and uncertainty, plus historical cases. If required inputs or professional review are missing, it returns needs-evidence.
Source text
SKILL.md · MIT
--- name: experience-score-calibrator description: "Calibrate explainable experience dimensions against reviewed evidence before exposing an aggregate score." --- # Experience Score Calibrator Keep scoring grounded in evidence coverage, uncertainty and human outcomes. ## Required evidence - dimension results - human judgments - coverage and uncertainty - historical cases ## Decisions - Calibrate dimensions before totals - Show confidence and missing evidence - Avoid false precision ## Boundary Do not publish score jumps without a validated rubric and sufficient cases. Return structured evidence, findings, proposed updates and limitations. Use `needs-evidence` when required inputs or professional review are missing. Never invent an Experience Score or promote a project observation into a universal rule.
© ShuaiMXu · MIT · ShuaiMXu/happyhands