By flatrick
ModelDev Toolkit (MDT) for Claude Code: a Node-only toolkit of agents, skills, hooks, commands, and rules for stable software workflows
Review instinct status and run the supported ai-learning workflow
Incrementally fix build and type errors with minimal, safe changes.
Create or verify a checkpoint in your workflow.
Comprehensive security and quality review of uncommitted changes:
Verify the working tree is clean enough to commit, run targeted checks, then draft and execute a conventional commit.
Expert Bash/shell script reviewer specializing in POSIX compatibility, security hardening, error handling, and cross-platform shell scripting. Use for all Bash/shell script changes.
Software architecture specialist for system design, scalability, and technical decision-making. Use PROACTIVELY when planning new features, refactoring large systems, or making architectural decisions.
Build and TypeScript error resolution specialist. Use PROACTIVELY when build fails or type errors occur. Fixes build/type errors only with minimal diffs, no architectural edits. Focuses on getting the build green quickly.
Personal communication chief of staff that triages email, Slack, LINE, and Messenger. Classifies messages into 4 tiers (skip/info_only/meeting_info/action_required), generates draft replies, and enforces post-send follow-through via hooks. Use when managing multi-channel communication workflows.
Expert code review specialist. Proactively reviews code for quality, security, and maintainability. Use immediately after writing or modifying code. MUST BE USED for all code changes.
Instinct-based learning system that adapts to the environment: hooks-passive observation for Claude/Cursor, explicit capture for hook-free tools. Single unified surface for all tools.
REST API design patterns including resource naming, status codes, pagination, filtering, error responses, versioning, and rate limiting for production APIs.
Patterns and architectures for autonomous MDT loops — from simple sequential pipelines to RFC-driven multi-agent DAG systems.
Backend architecture patterns, API design, database optimization, and server-side best practices for Node.js, Express, and Next.js API routes.
Safe, portable, and idiomatic Bash/shell scripting patterns including error handling, security hardening, argument parsing, testing with bats-core, and cross-platform techniques.
Matches all tools
Hooks run on every tool call, not just specific ones
Executes bash commands
Hook triggers when Bash tool is used
ModelDev Toolkit (MDT) is a Node-first workflow toolkit for Claude Code, Cursor, and Codex. It ships agents, skills, commands, rules, and install adapters behind one public CLI: mdt.
git clone https://github.com/flatrick/modeldev-toolkit.git
cd modeldev-toolkit
npm install
npx mdt install typescript
npx mdt verify tool-setups
Tool-specific installs:
npx mdt install --tool cursor typescript
npx mdt install --tool codex typescript ai-learning
npx mdt install list
Cursor IDE note:
~/.cursor/mdt bridge materialize --tool cursor --surface rulesWindows + Cursor hooks: On Windows, Cursor may pass large hook payloads via spawn arguments instead of stdin. That can trigger spawn ENAMETOOLONG before the hook script runs—so the script cannot fix it. Affected hooks include beforeReadFile, afterFileEdit, beforeSubmitPrompt, and sessionEnd. See Cursor hooks limitation for workarounds (remove the hook or use payload-via-temp-file when Cursor supports it).
Root entrypoints:
Current-state docs:
ECC comparison:
mdt ... as the official public entrypoint.docs/ as the source of truth for current install and capability behavior.agents/ shared agent promptscommands/ shared markdown commandsskills/ shared skill definitionsrules/ shared rule packscursor-template/, codex-template/, claude-template/ tool install-source treesdocs/ current docs, manual verification, plans, and ECC comparisonscripts/ Node.js runtime, install, and CI helpersRepository checks:
npm run lint
npm test
npm run test:verbose
npm test now keeps console output focused on failed/skipped items and writes detailed JSONL artifacts under .artifacts/logs/test-runs/.
Tool setup checks:
mdt verify tool-setups
Dev-install maintainer checks (--dev only):
mdt dev smoke tool-setups
mdt dev smoke workflows --tool claude
mdt dev smoke workflows --tool cursor
mdt dev smoke workflows --tool codex
MIT. Original upstream work by @affaan-m, maintained independently here.
Modifies files
Hook triggers on file write and edit operations
Modifies files
Hook triggers on file write and edit operations
Uses power tools
Uses Bash, Write, or Edit tools
Uses power tools
Uses Bash, Write, or Edit tools
No model invocation
Executes directly as bash, bypassing the AI model
No model invocation
Executes directly as bash, bypassing the AI model
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
npx claudepluginhub flatrick/mdtClaude Code workflow patterns: prompting, CLAUDE.md maintenance, multi-agent orchestration
Complete collection of battle-tested Claude Code configs from an Anthropic hackathon winner - agents, skills, hooks, and rules evolved over 10+ months of intensive daily use
Interactive setup wizard for configuring any repository with Claude Code best practices, based on Boris Cherny's workflow
Comprehensive skills library for Claude Code: planning, design, TDD, debugging, collaboration patterns, and proven techniques
Battle-tested skills, agents, and commands for Claude Code — TDD workflows, systematic debugging, code review, and parallel task execution.
Harness-native ECC plugin for engineering teams - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, MCP conventions, and operator workflows for Claude Code plus adjacent agent harnesses
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 claim