By benshapyro
Enforce code quality standards across full-stack JavaScript and Python projects with automated code review, testing, git workflow management, and structured project scaffolding from concept to documentation.
You are in **Backlog Mode** - documenting bugs and improvements WITHOUT implementing them.
Discover and specify a new software project from scratch
Learn about Claude Code, the devkit, commands, skills, and workflows
Plan a new feature or task with proper structure before implementation.
Save research findings as reusable knowledge documents
Reviews code for adherence to coding standards, best practices, and style guidelines. PROACTIVELY use after completing any significant code changes (new features, refactors, bug fixes). Auto-invoke when code is ready for review before committing.
Systematically analyzes errors, stack traces, and logs to identify root causes. PROACTIVELY use when user reports bugs, errors, crashes, or unexpected behavior. Auto-invoke when stack traces or error messages appear in conversation.
Researches official documentation, latest best practices, and current tech stack information. PROACTIVELY use when evaluating libraries, checking API usage, or when implementation involves unfamiliar frameworks. Auto-invoke before using any library/API features from training data.
Assists with Git operations following branching strategy and commit conventions. PROACTIVELY use when creating branches, making commits, resolving conflicts, or managing git workflows. Auto-invoke for any git-related questions or operations.
Analyzes and optimizes code performance. PROACTIVELY use when user mentions slow, laggy, timeout, memory issues, or optimization needs. Auto-invoke when performance-critical code is being written (loops, data processing, API calls).
Provides REST and GraphQL API design patterns for Node.js, Flask, and FastAPI. Use when designing endpoints, request/response structures, API architecture, pagination, authentication, rate limiting, or when working in /api/ or /routes/ directories.
Formats code according to Ben's style guidelines for TypeScript, Python, and general best practices. Use when formatting code, fixing linting issues, checking naming conventions, organizing imports, or when user mentions formatting, style, linting, Prettier, Black, or ESLint.
Knowledge base for the Cadre DevKit. Use when answering questions about the devkit structure, commands, skills, hooks, agents, or workflows.
Generates README files, API documentation, and inline code comments following best practices. Use when creating project documentation, writing READMEs, documenting APIs, or explaining complex code.
Provides battle-tested error handling patterns for TypeScript and Python. Use when implementing error handling, creating try/catch blocks, or handling exceptions.
Executes bash commands
Hook triggers when Bash tool is used
Modifies files
Hook triggers on file write and edit operations
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Uses power tools
Uses Bash, Write, or Edit tools
Uses power tools
Uses Bash, Write, or Edit tools
Runs pre-commands
Contains inline bash commands via ! syntax
Runs pre-commands
Contains inline bash commands via ! syntax
Turn Claude Code from a helpful intern into a reliable senior engineer.
Claude Code is Anthropic's official CLI for AI-assisted coding. You run it in your terminal, and Claude can read your files, write code, run commands, and help you build software.
It's powerful. It's also... unpredictable.
AI coding assistants have a reliability problem:
| Issue | What Happens |
|---|---|
| Hallucination | "Tests pass!" (they don't) |
| Over-engineering | 50 lines of defensive code for a 3-line function |
| AI slop | Comments everywhere, any casts, unnecessary try/catch |
| Context loss | Re-explores the same code every session |
| Dangerous commands | rm -rf / is just a typo away |
| No planning | Jumps straight to code without understanding |
| Inconsistent quality | Great on Monday, chaos on Tuesday |
You end up babysitting the AI instead of shipping code.
This devkit adds a quality and safety layer to Claude Code:
It's not a collection of prompts. It's an integrated system.
Start Claude Code:
claude
Then inside Claude Code, add the marketplace and install the plugin:
/plugin marketplace add github.com/benshapyro/cadre-devkit-claude
/plugin install cadre-devkit-claude@cadre-devkit
For local development, use the file path:
/plugin marketplace add /Users/bshap/Projects/cadre-internal/devkits/cadre-devkit-claude
/plugin install cadre-devkit-claude@cadre-devkit
git clone https://github.com/benshapyro/cadre-devkit-claude.git
cd cadre-devkit-claude
./install.sh
# Clone
git clone https://github.com/benshapyro/cadre-devkit-claude.git
# Copy components
cp -r cadre-devkit-claude/commands/* ~/.claude/commands/
cp -r cadre-devkit-claude/skills/* ~/.claude/skills/
cp -r cadre-devkit-claude/agents/* ~/.claude/agents/
cp -r cadre-devkit-claude/hooks/* ~/.claude/hooks/
# Make hooks executable
chmod +x ~/.claude/hooks/**/*.py
chmod +x ~/.claude/hooks/**/*.sh
Add to ~/.claude/settings.json:
{
"hooks": {
"PreToolUse": [
{
"matcher": "Bash",
"hooks": [{ "type": "command", "command": "~/.claude/hooks/security/dangerous-command-blocker.py" }]
},
{
"matcher": "Edit|Write|Read",
"hooks": [{ "type": "command", "command": "~/.claude/hooks/security/sensitive-file-guard.py" }]
}
],
"PostToolUse": [
{
"matcher": "Edit|Write",
"hooks": [
{ "type": "command", "command": "~/.claude/hooks/formatting/auto-format.py" },
{ "type": "command", "command": "~/.claude/hooks/testing/test-on-change.py" }
]
}
]
}
}
After installing via plugin system, restart Claude Code to load the plugin. Then verify:
/help # Should show /learn, /plan, /research, etc.
/learn # Should show welcome message
Can you run rm -rf / # Should be blocked by security hook
If you used the install script or manual installation, hooks are configured automatically in your ~/.claude/settings.json.
npx claudepluginhub benshapyro/cadre-devkit-claudeHarness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Comprehensive feature development workflow with specialized agents for codebase exploration, architecture design, and quality review
Tools to maintain and improve CLAUDE.md files - audit quality, capture session learnings, and keep project memory current.
Access thousands of AI prompts and skills directly in your AI coding assistant. Search prompts, discover skills, save your own, and improve prompts with AI.
Develop, test, build, and deploy Godot 4.x games with Claude Code. Includes GdUnit4 testing, web/desktop exports, CI/CD pipelines, and deployment to Vercel/GitHub Pages/itch.io.