Modular Claude Code marketplace — 25 domain-focused plugins, natively orchestrated with zero forcing. MIT license.
npx claudepluginhub arnwaldn/atum-plugins-collectionFoundation plugin for the ATUM Plugins Collection. Provides safety hooks (secret scanner, git guard, anti-rationalization, loop detector, pre-completion gate, file guard, lock file protector), common rules (coding style, security, testing, anti-hallucination, decision framework), baseline commands (commit, verify, health, checkpoint), and settings permissions. Required by all other ATUM plugins.
High-level development workflows for the ATUM ecosystem. Provides autopilot (idea to deployed product in 6 phases), feature development pipelines, scaffolding, project interview, TDD cycles, deployment orchestration, and planning agents (planner, architect, code-architect, codebase-pattern-finder, code-explorer, fresh-executor). Depends on atum-core.
Multi-model orchestration via OpenRouter. Provides a custom MCP server for consulting free and paid models (jury, debate, expert modes), two orchestrator agents (legion-commander, legion-strategist), and the /legion command for strategic consultations. Supports cost tracking, budget guardrails, and dynamic model selection across 100+ providers. Perfect for architecture decisions, code reviews, and multi-perspective analysis. Depends on atum-core.
Cross-session memory and learning layer for Claude Code. Integrates with claude-mem for persistent observations, provides /recall and /remember commands for quick memory access, continuous-learning v2 for instinct refinement, and session lifecycle hooks. Enables Claude to become a true long-term collaborator who remembers context across sessions. Depends on atum-core.
Multi-language code review specialists. Includes reviewers for TypeScript, Python, Go, Rust, C++, Java, Kotlin, Flutter, database (PostgreSQL), security (OWASP Top 10), and architecture. Also provides silent-failure-hunter, pr-test-analyzer, comment-analyzer, type-design-analyzer, and skill-reviewer agents. Commands: /code-review, /review-pr, /security-audit, /test-coverage. Depends on atum-core.
Build error resolution specialists for 7 language stacks. Fixes compilation errors, linker issues, type check failures, and dependency problems with minimal-diff changes. Includes build-error-resolver (generic), cpp-build-resolver, go-build-resolver, java-build-resolver, kotlin-build-resolver, pytorch-build-resolver, rust-build-resolver, error-detective, and error-analyzer. Commands: /build-fix, /cpp-build, /go-build, /kotlin-build, /rust-build, /fix-issue. Depends on atum-core.
Modern web development stack. TypeScript + React + Next.js + Vue/Nuxt patterns, frontend design systems, accessibility (WCAG 2.1 AA), E2E testing (Playwright), D3.js data visualization, Turbopack optimization, high-performance browser techniques, and component design. Skills for frontend patterns, UI/UX Pro Max, design handoff, and implementation from Figma. Depends on atum-core.
Backend development stacks for multiple languages. Python (Django, FastAPI), Java (Spring Boot, JPA), Go (idiomatic patterns, testing), Rust (ownership, lifetimes), C++ (modern idioms, testing), PHP (Laravel), Perl (modern patterns). Includes common architecture patterns: clean architecture, domain-driven design, architecture decision records, API design, resilience, system design, DDIA. Depends on atum-core.
Mobile development stacks. Swift/iOS with SwiftUI patterns, Swift 6.2 concurrency, actor persistence, protocol-based DI, Foundation Models on-device, Liquid Glass design for iOS 26. Kotlin/Android with coroutines, Flow, testing patterns. Flutter/Dart with widget best practices and clean architecture. Expo/React Native. Tauri for cross-platform desktop-mobile. Includes Sentry setup for iOS Swift. Depends on atum-core.
No-code automation mastery. Design, audit, and optimize Make.com scenarios (blueprints from natural language), Airtable bases (relational schemas, optimization), and Notion workspaces (wikis, databases, automations). Includes no-code-maestro skills, Pinecone Assistant integration, and commands for daily status, audit-airtable, audit-notion, diagnose-make. Designed for ATUM Agency's day-to-day client work. Depends on atum-core.
AI/ML and LLM engineering workflows. HuggingFace ecosystem (CLI, datasets, evaluation, jobs, model training, paper publishing, tool building, Trackio). PyTorch patterns. RAG architecture. Claude API best practices. Cost-aware LLM pipelines with intelligent model routing. AI regression testing. Regex-vs-LLM decision framework. Evaluation harness. Data engineering and ML engineering agents. Depends on atum-core.
Regulatory compliance and security audits. EU AI Act (Article 15 traceability via ATUM Audit integration), open-source license compliance (detection and audit), supply chain risk auditing, security scanning. Includes compliance-expert and security-expert agents, /atum-audit, /compliance, and /security-audit commands. Core differentiator for European enterprises facing regulatory requirements. Depends on atum-core.
Game development specialists for major engines and frameworks. Includes agents for Unity (C#), Godot (GDScript), Phaser (TypeScript web games), Three.js (3D web), and a generalist game-architect. Covers design patterns, performance optimization, cross-platform deployment, and game-specific architecture concerns. Depends on atum-core.
Cryptographic file integrity monitoring with OWL/RDF ontology and EU AI Act (Reg. 2024/1689) compliance. 15 MCP tools for file hashing, SPARQL queries, SHACL validation, Annex IV documentation, incident tracking, and compliance reporting. 3 automatic hooks (SessionStart, PostWrite, PostCommit). Strategic compliance backbone for European enterprises.
Complete dev project preparation — from qualification to pre-dev gate. Adaptive S/M/L/XL workflow with 82 questions, 8 deliverables, and 2 bridge files (CLAUDE.md + implementation-roadmap.md). 4 specialized agents, 7 templates, adaptation matrix.
Lightweight gateway to 164 scaffold templates, 18 compliance checklists, and 10 reference catalogs. Fetches data on-demand from the upstream arnwaldn/project-templates GitHub repo — zero duplication, always up-to-date.
Odoo v19 specialists — 12 domain-expert agents (ORM, accounting, manufacturing, retail/POS, healthcare, hospitality, education, events, field service, real estate, food & beverage, third-party integrations) plus 4 commands (/odoo, /odoo-audit, /odoo-migrate, /odoo-module). Includes an odoo-knowledge skill aggregating 15 reference guides (v19 core, security, performance, testing, deployment, OCA standards, Owl frontend, SaaS, multi-company, API, migration, ITPP patterns, MCP integration). The only Odoo-focused plugin in the Claude Code marketplace. Depends on atum-core.
Algorithmic trading and MetaTrader 5 specialist plugin — 1 expert agent (mql5-expert) covering Expert Advisors, custom indicators, scripts, services and libraries, the /mql5 command for scaffolding trading programs, and a trading-knowledge skill aggregating MQL5 complete guide, proven trading strategies, and a trading brain architecture reference. The only dedicated algorithmic-trading plugin in the Claude Code marketplace. Depends on atum-core.
Content management systems and e-commerce specialists — domain experts for WordPress/WooCommerce, Shopify (Liquid + Hydrogen), and headless CMS (Sanity, Strapi, Payload, Webflow, Ghost). Covers theme/app development, Liquid/Gutenberg/GROQ/Handlebars templating, Storefront and Admin APIs, custom plugins and extensions, content modeling, live preview, i18n, and commerce-specific workflows (products, carts, checkout, metafields, webhooks, subscriptions). The only plugin in the ATUM marketplace dedicated to CMS and commerce platforms that real client projects actually require. Depends on atum-core.
Persistent cross-session memory, intelligent context routing, and reflective journaling for the ATUM ecosystem. Gives Claude a real brain via an MCP server with 8 tools: remember, recall, project context, journal, reflect, suggest context, and status dashboard.
Live web search (Tavily) and URL extraction. Optimized for AI agents to retrieve structured, LLM-ready data from the web (search, extract, map, crawl). Zero dependencies.
Native OS notifications and Discord/Slack webhooks. Informs the user about session end or task completion. Features an automatic trigger on session stop.
Claude Code marketplace entries for the plugin-safe Antigravity Awesome Skills library and its compatible editorial bundles.
Directory of popular Claude Code extensions including development tools, productivity plugins, and MCP integrations
Curated collection of 129 specialized Claude Code subagents organized into 10 focused categories