ai-toolkit

Create Architecture Decision Records through interactive conversation

Get implementation guidance for a phase without automated code generation - user implements manually

Unified branch operations with git-workflow enforcement

Check if CHANGELOG is up to date and update as necessary

Create a proper git commit with quality checks and conventional message

Complete any issue type per its workflow requirements

Unified documentation management - generate, validate, sync, update, and analyze documentation

Execute specific implementation phases from task plans with test-first enforcement

Create standalone work items (TASK, BUG, or SPIKE) with AI-assisted type detection

Add AI-suggested comments to external Jira issues based on work context

Create Jira epics through natural language conversation (requires Jira integration)

Import Jira issue for local work with PLAN.md creation

Promote local exploration issue to Jira for team visibility

Create PLAN.md file with phase-based breakdown for tasks, bugs, and spikes

Fill and improve project brief through gap-driven conversation

Enhanced project status dashboard with intelligent context analysis

Comprehensive quality assessment with multi-agent coordination

Silently refresh AI context by reading project configuration, guidelines, and recent commits

Release new version following semantic versioning guidelines

Deep research on a topic, creating persistent documentation for future reference

Step back, reflect on current work, validate direction, and assess alignment with plan and architecture

Create local feature specifications through natural language conversation

Sync project state after manual changes via git diff analysis

Initialize project with ai-toolkit structure - minimal questions, fast setup

5-step debugging loop for unexpected issues during implementation or standalone bug fixes

Create and iterate on HTML UI mockups with parallel design exploration

Add timestamped work log entries to track manual changes and communicate with AI

**AUTOMATICALLY INVOKED for AI and LLm related design and implementation.** Expert AI researcher and practitioner with deep knowledge of Large Language Models, their architectures, capabilities, and practical applications. Use for AI/ML technology selection, model comparison, context management strategies, prompt engineering, RAG systems, and emerging AI trends. Provides authoritative guidance on AI integration, implementation patterns, and optimization.

**AUTOMATICALLY INVOKED for API design and implementation tasks.** Design REST APIs, GraphQL schemas, service interfaces, and data validation patterns. **Auto-invoked when** designing endpoints, defining API contracts, or planning service integrations. Focus on developer experience, consistency, and robust error handling.

**AUTOMATICALLY INVOKED for AWS cloud architecture and implementation.** Expert AWS Solutions Architect with deep knowledge of AWS services, architectures, and best practices. Use for AWS service selection, architecture design, cost optimization, security best practices, and implementation guidance. Provides authoritative guidance on AWS-specific solutions and patterns.

**AUTOMATICALLY INVOKED for Azure cloud architecture and implementation.** Expert Azure Solutions Architect with deep knowledge of Microsoft Azure services, architectures, and best practices. Use for Azure service selection, architecture design, cost optimization, security best practices, and implementation guidance. Provides authoritative guidance on Azure-specific solutions and patterns.

**AUTOMATICALLY INVOKED for server-side implementation tasks.** Expert-level backend specialist for implementing robust, scalable server-side applications. **Use immediately when** building APIs, implementing business logic, setting up authentication, real-time features, or background processing.

AUTOMATICALLY INVOKED for /project-brief commands. Strategic brief specialist focused on product strategy, market positioning, and business model design. . Conducts interactive discovery process with structured questioning to gather all project brief elements before generating documents.

**MANDATORY for all /plan reviews.** Proactively reviews system-wide design decisions, architectural planning, and cross-cutting concerns. **Auto-invoked during planning phase** to validate architecture before implementation. Use for complex features requiring architectural design, technology decisions, or changes affecting multiple system components.

**Use PROACTIVELY after code implementation.** Thorough code reviews focusing on quality, maintainability, security, and adherence to project standards. **Auto-invoked after** significant code changes to ensure quality before commit. Reviews code for best practices, potential issues, performance implications, and architectural alignment.

**AUTOMATICALLY INVOKED for all database work.** Use for schema design, query optimization, migrations, performance tuning, and data architecture. **Use immediately when** designing databases, optimizing queries, creating migrations, or addressing performance issues. Focus on scalability, data integrity, and optimal performance.

Infrastructure specialist and deployment automation expert focused on creating robust, scalable, and secure development and production environments. Auto-invoked for infrastructure setup, deployment automation, and CI/CD pipeline issues.

**AUTOMATICALLY INVOKED for UI/UX implementation tasks.** Expert-level frontend specialist for component creation, responsive design, and user interface development. **Use immediately when** building components, implementing designs, optimizing frontend performance, or handling accessibility requirements.

**AUTOMATICALLY INVOKED for Google Cloud architecture and implementation.** Expert Google Cloud Solutions Architect with deep knowledge of GCP services, architectures, and best practices. Use for GCP service selection, architecture design, cost optimization, security best practices, and implementation guidance. Provides authoritative guidance on Google Cloud-specific solutions and patterns.

Version upgrades, framework migrations, and dependency updates specialist. AUTOMATICALLY INVOKED for safe migrations with compatibility assessment, incremental modernization strategies, and comprehensive rollback planning.

Performance analysis and optimization specialist. Auto-invoked for performance bottlenecks, slow queries, optimization requests, and scalability concerns.

PROACTIVELY orchestrates multiple specialized agents for complex, multi-domain tasks AND serves as a general-purpose agent when no specialist is suitable. Use for feature development, system-wide changes, multi-domain tasks, or general research and analysis. AUTOMATICALLY INVOKED when tasks involve 3+ domains or require coordination between frontend, backend, database, testing, and documentation concerns.

Code improvement, cleanup, and technical debt reduction specialist. Auto-invoked for refactoring requests, code quality improvements, and technical debt reduction.

Deep research specialist that reads extensively and distills to essentials. Auto-invoked when agents need external knowledge. Creates persistent research documents for reuse. "I read 30 blog posts, here are the 3 that matter."

**AUTOMATICALLY INVOKED for security-relevant tasks.** Proactively reviews authentication, authorization, data protection, and compliance. **Use immediately when** implementing auth systems, handling sensitive data, or making security-critical changes. Focus on OWASP Top 10, compliance standards, and secure architecture validation.

Comprehensive documentation specialist handling creation, maintenance, and synchronization. AUTOMATICALLY INVOKED when code changes affect documentation or when new documentation is needed. Provides bidirectional sync between code and documentation, automatic generation capabilities, and ensures all documentation follows project standards.

Comprehensive test creation, test strategy development, and test suite maintenance. Use PROACTIVELY for TDD/BDD workflows, creating test suites for new features, test automation, and maintaining test quality. AUTOMATICALLY INVOKED when test failures are detected to analyze and resolve issues.

Use PROACTIVELY when user requests design work, mockups, color schemes, accessibility guidance, or design system decisions. Specializes in user experience, visual design, accessibility compliance (WCAG), and design systems. Coordinates with frontend-specialist for implementation.