Skip to main content
skill CX Codex

ai-eval-playbook — Claude Code skill for AI evaluation

ai-eval-playbook is a Claude Code skill for assessing AI tools, chatbots, and GenAI products. It comes into play when a team needs evaluation measures or a review of an AI tool proposal.

no license 118 installs ★ 3 updated 2026-10-23

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add gasserane/personal-skills --skill ai-eval-playbook -a claude-code
Open on GitHub

What ai-eval-playbook does

The skill frames product evaluation across four levels: model, product, user, and impact. Its scope includes minimum viable evaluations, golden datasets, LLM-as-judge methods, engagement measures, and guardrail metrics.

Before giving a substantive answer, it requires fetching the current playbook from https://eval.playbook.org.ai/. The fetch options include WebFetch and ctx_fetch_and_index, with claude-in-chrome as a fallback. It does not cover programme or MEL evaluation design, or AI used as an evaluator’s working tool.

What it helps with

  • choose an evaluation level for an AI product
  • select measures for a chatbot evaluation
  • appraise an AI tool proposal

Example prompts for the agent

  • How should we evaluate this AI tool?
  • Which measures should we track for this chatbot?
  • Review this partner’s AI tool proposal and outline an evaluation.

How to install ai-eval-playbook

1Skills CLI (Vercel)

npx skills add gasserane/personal-skills --skill ai-eval-playbook -a claude-code

2Manually: put the files in place

~/.claude/skills/ai-eval-playbook/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/ai-eval-playbook/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about ai-eval-playbook

How do I install ai-eval-playbook in Claude Code?

Run npx skills add gasserane/personal-skills --skill ai-eval-playbook -a claude-code. The skill file is at ~/.claude/skills/ai-eval-playbook/SKILL.md.

What does ai-eval-playbook do?

It helps assess AI models and products across four levels, from model performance to impact. It covers evaluation measures, test datasets, and safeguards.

What does ai-eval-playbook need to work?

Substantive answers require fetching the current guide from https://eval.playbook.org.ai/. The skill lists WebFetch, ctx_fetch_and_index, and claude-in-chrome as ways to retrieve it.

Similar skills

All skills for AI and ML →

From the same set44

The whole set →
linkedin-field-note linkedin-field-note — a skill for Claude Code runs an end-to-end editorial pipeline for personal AI Field Notes on LinkedIn, covering topic triage, dictation, kit generation, and posting. CX Codex +1 219installs donor-proposal-scoring donor-proposal-scoring — a skill for Claude Code scores grant proposals against a donor’s published criteria. It can check thresholds or compare its assessment with the donor’s evaluation report. CX Codex +1 182installs office-repair office-repair fixes complex Excel workbooks in place via COM without dropping slicers, or promotes hand-edited Office files to canonical sources. CX Codex +1 144installs office-review-pass office-review-pass runs a .docx through a review cycle: reads other people's comments, writes findings back as margin notes, and applies Word tracked changes a counterparty can accept or reject. CX Codex +1 144installs selection-toolkit selection-toolkit is a Claude Code skill that builds Excel workbooks for selection and multi-criteria appraisal, with weighted scoring based on published criteria. CX Codex +1 135installs contract-delivery-kickoff contract-delivery-kickoff turns a signed supplier contract into a first-session pack: a workbook that totals days against the contracted ceiling, an internal brief, a supplier agenda and a note template. CX Codex +1 134installs literature The literature Claude Code skill searches scholarly works through OpenAlex. It runs to find sources, check whether a question has been studied, or map key papers. CX Codex +1 130installs media-to-roadmap The media-to-roadmap Claude Code skill turns external tech and AI material into a system-improvement roadmap. It starts from a link or a request to assess a video, podcast, article, or thread. CX Codex +1 123installs

Project sponsors

Companies that keep this analytics open