Skillget

claude-code-setup-evaluator

skill
Guvenlik Denetimi
Basarisiz
Health Gecti
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Community trust — 10 GitHub stars
Code Basarisiz
  • rm -rf — Recursive force deletion command in .ai-workspace/tests/test-hooks.sh
Permissions Gecti
  • Permissions — No dangerous permissions requested

Bu listing icin henuz AI raporu yok.

SUMMARY

A meta-repository for Claude Code users that includes workspace setup (skills, commands, hooks) and an evaluator for your Claude Code configuration

README.md

Claude Code Setup Evaluator

A workspace for Claude Code users who work across multiple repositories. Clone your repos, get shared skills and commands out of the box, add your own — and use /evaluate-setup to make sure your setup is actually helping.

How It Works

This is a meta-workspace. You clone your project repositories into the repositories/ folder and run Claude Code from the workspace root. Claude automatically gets access to all skills and commands. Your repos stay independent — they push to their own remotes, the workspace never touches them.

claude-code-setup-evaluator/
  repositories/           # Clone your repos here (gitignored — yours, not shared)
    my-api/
    frontend/
    data-pipeline/
  skills/                 # AI skills — activate automatically when relevant
  commands/               # Slash commands — you trigger with /command
  .claude/                # Claude Code hooks and config
  scripts/evaluate-setup/ # The evaluator engine

Quick Start

git clone https://github.com/redhat-community-ai-tools/claude-code-setup-evaluator.git
cd claude-code-setup-evaluator
uv sync
uv run .ai-workspace/scripts/setup.py

# Clone your repos
cd repositories/
git clone <your-repo-url>
cd ..

# Start working
claude

Tell Claude which repo to focus on — "I'm working on my-api" or type /focus to pick from a list. Skills activate automatically. Commands are there when you need them.

What You Get

Skills (activate automatically)

Skills are knowledge that Claude carries in the background. You don't trigger them — they activate when relevant to what you're doing.

Skill What it does
security-check Detects credential leaks, insecure patterns, LLM-specific risks
python-conventions Dotenv conventions, API client rules, TDD workflow
data-pipeline-patterns Pipeline stage design, validation, circuit breakers
verification-loop Powers /verify, /review, /quality-gate
brainstorming Design exploration before implementation
writing-plans Creates implementation plans from approved specs

Commands (you trigger these)

Command When to use What it does
/evaluate-setup Anytime Evaluate your entire Claude Code setup (see below)
/verify After coding Check types, lint, tests
/review Before pushing Code review with security checks
/quality-gate Before git push Tests + secret scan
/refactor-safe After review Refactor without changing public API
/test-coverage Adding tests Find untested code, generate tests
/diff-explain Reviewing changes Explain by intent, not file count
/explain-code Unfamiliar code Layered explanation scaled to complexity
/explain-simple Non-technical audience Plain language, no jargon
/prompt-test Editing LLM prompts Test against samples, catch regressions
/ai-engineer-review Architecture check Brutally honest review
/architecture-docs Documentation Architecture docs with Mermaid diagrams
/env-check Something broke Validate environment setup
/recap End of session Summarize for standup
/focus Switch repos Pick repos from a list
/toolkit Discovery See everything available

Hooks (automatic safety nets)

Hook What it does
Session start Reports git status of all your repos
Secret scan Blocks git commit/push if API keys detected in tracked files
Skill suggestion Reminds Claude which skills are relevant to files you're editing

Adding Your Own Skills and Commands

You can extend the workspace with your own skills and commands:

  • Add a skill — Create a folder in skills/ with a SKILL.md file. Claude picks it up automatically.
  • Add a command — Create a folder in commands/ with a command.md file. It becomes a /command you can type.

After adding, run /evaluate-setup to check that your new skill or command meets quality standards.

The Evaluator (/evaluate-setup)

The built-in evaluation command checks whether your skills, commands, CLAUDE.md, and hooks are actually helping — or wasting context and making Claude worse.

Layer 1 — Static Analysis. A rule engine with 15 rules scans for mechanical issues: missing descriptions, broken references, duplicate content, prompt injection patterns, token budget violations.

Layer 2 — AI Review. Claude scores each item on a structured rubric (5 dimensions, 1-5 scale) and suggests cross-type optimizations — e.g., "this skill should be a hook" or "move this from CLAUDE.md to a skill."

Layer 3 — A/B Testing (optional). Tests whether skills actually change Claude's behavior by running tasks with and without each skill on your actual repositories, then using Gemini to judge the difference with redundancy-first scoring.

You: /evaluate-setup

Claude: What do you want to evaluate?
        > Everything / Skills only / A specific item

        ## Evaluation Summary
        Your setup is solid. Found 1 issue that needs attention.
        Reviewed 6 skills, 17 commands, CLAUDE.md. Total: 8,234 tokens (4%).

        Suggestions:
          1. Remove clean-code-guide skill (100% redundant with Claude defaults)

Note: skills/clean-code-guide and commands/check-code are intentionally low-quality examples included for testing the evaluator. Run /evaluate-setup to see how they get flagged.

Cursor Support

The main branch is Claude Code only. For Claude Code + Cursor support, see the multi-tool branch.

Learn More

  • GUIDE.md — Full reference for all skills, commands, and hooks
  • docs/spec.md — Full evaluator specification

Credits

Built on top of ai-workspace-template by Michael Yochpaz.

Yorumlar (0)

Sonuc bulunamadi