Skip to content

ROADMAP.md

Latest commit

 

History

History
246 lines (186 loc) · 7.62 KB

ROADMAP.md

File metadata and controls

246 lines (186 loc) · 7.62 KB

AIWB Development Roadmap

This document outlines the development direction and milestones for the AIWB project. It reflects both completed work and future plans.

Last Updated: November 2025 (v2.0.1)

Philosophy

Our goal is to create the most powerful, flexible, and user-friendly command-line AI development environment. We believe in:

  • AI-Native Workflow: Integrating AI at every step, from concept to code.
  • Hybrid Intelligence: Leveraging the strengths of multiple AI models through Generator-Verifier loops.
  • User Empowerment: Giving developers granular control, cost transparency, and clear insights.

✅ Phase 1: Foundation & Refinement (COMPLETED - v2.0)

Objective: Solidify the core system, ensure stability, and streamline the developer experience.

Completed Features

  • ✅ Code Consolidation & Optimization:

    • Refactored into modular library system (lib/common.sh, lib/api.sh, lib/modes.sh, etc.)
    • Consistent error handling across all modules
    • Centralized configuration management
    • Eliminated redundant code paths

  • ✅ Robust Installer & Setup:

    • Fully automated install.sh script
    • Dependency checking (bash, curl, jq, gum)
    • Automatic workspace initialization
    • PATH configuration with ~/.local/bin

  • ✅ Cross-Platform Compatibility:

    • Full support for Linux, macOS, and Android (Termux)
    • Platform-specific adaptations (safe_read, clipboard, paths)
    • CRLF auto-fix for Termux
    • Fallback UI when gum not available

  • ✅ Core Documentation:

    • Comprehensive README.md and QUICKSTART.md
    • Detailed USAGE.md guide
    • OVERVIEW.md with design philosophy
    • WORKFLOW-GUIDE.md with patterns
    • Platform-specific guides (TERMUX_INSTALL.md)

✅ Phase 2: Feature Expansion & Model Agnosticism (COMPLETED - v2.0)

Objective: Broaden AIWB's capabilities and support a wider range of AI models.

Completed Features

  • ✅ Modular AI System:

    • Unified API layer (lib/api.sh) supporting multiple providers
    • Plugin architecture for easy provider addition
    • Integrated providers:
      • ✅ Google Gemini (gemini-2.5-flash, gemini-2.0-flash-lite)
      • ✅ Anthropic Claude (claude-3.5-sonnet, claude-3.5-haiku)
      • ✅ OpenAI (gpt-4o, gpt-3.5-turbo)
      • ✅ Groq (llama-3.3-70b-versatile, mixtral-8x7b, vision models)
      • ✅ xAI/Grok (grok-beta, grok-2, grok-code-fast-1)
      • ✅ Ollama (local model support)

  • ✅ Interactive TUI:

    • Beautiful terminal UI using gum with fallback support
    • Interactive menus for model/provider selection
    • File browser with multi-storage support
    • Progress indicators and status displays
    • Colored output and formatting

  • ✅ Mode-Based Workflows:

    • /make mode - Generate from scratch
    • /tweak mode - Modify existing code
    • /debug mode - Find and fix bugs
    • Structured workflow with prompt, uploads, verification
    • Cost estimation before execution

  • ✅ Cost Management:

    • Pre-flight cost estimation
    • Real-time usage tracking (usage.jsonl)
    • Per-provider spending breakdown
    • Cost-aware model selection

  • ✅ Security:

    • Age-encrypted API key storage
    • Secure key management interface
    • Proper file permissions (600 for sensitive files)

🚧 Phase 3: Enhanced Workflows & Intelligence (In Progress - v2.1)

Objective: Add more intelligent workflows, better context handling, and enhanced user experience.

Planned Features

  • Enhanced Context Management:

    • 🔲 Automatic context relevance detection
    • 🔲 Smart file filtering based on task type
    • 🔲 Context size optimization
    • 🔲 Incremental context loading

  • Advanced Mode Features:

    • 🔲 /plan mode - Planning and architecture design
    • 🔲 /refine mode - Multi-iteration autonomous refinement loop
    • 🔲 Mode-specific templates and presets
    • 🔲 Custom mode creation

  • Git Integration:

    • 🔲 AI-powered commit message generation from git diff
    • 🔲 Pre-commit code review with AI
    • 🔲 Branch analysis and merge conflict resolution
    • 🔲 Changelog generation

  • Improved UI/UX:

    • 🔲 Syntax highlighting in outputs
    • 🔲 Diff visualization for code changes
    • 🔲 Markdown rendering in terminal
    • 🔲 Progress bars for long operations
    • 🔲 Notification system for completion

  • Template System Enhancement:

    • 🔲 Template marketplace/sharing
    • 🔲 Dynamic template variables
    • 🔲 Template inheritance and composition
    • 🔲 Interactive template builder

🔮 Phase 4: Autonomous Agent Capabilities (Future - v3.0)

Objective: Transform AIWB into a semi-autonomous development partner with agentic capabilities.

Vision Features

  • Autonomous Refinement Loop:

    • Fully autonomous Generator-Verifier iterations
    • Intelligent stopping criteria (convergence detection)
    • Quality metrics and improvement tracking
    • Configurable iteration limits and goals

  • File System Agent:

    • AI can propose and execute file operations (with user confirmation)
    • Multi-file refactoring capabilities
    • Codebase-wide changes with safety checks
    • Automated testing after modifications

  • Test-Driven Generation (TDG):

    • Generate failing tests based on requirements
    • Generate code to make tests pass
    • Iterative test refinement
    • Coverage analysis and optimization

  • Advanced Prompt Engineering:

    • Version control for prompts (git-like)
    • Prompt templates with variables
    • A/B testing for prompt effectiveness
    • Community prompt library

  • Workflow Orchestration:

    • Multi-step workflow definitions
    • Conditional logic and branching
    • Parallel task execution
    • Workflow resumption and rollback

  • Learning System:

    • Learn from user corrections
    • Project-specific patterns and preferences
    • Custom style guides enforcement
    • Feedback loop for continuous improvement

🌟 Phase 5: Ecosystem & Integrations (Future - v3.5+)

Objective: Build an ecosystem around AIWB with integrations and community features.

Future Ideas

  • Editor Integrations:

    • VS Code extension
    • Vim/Neovim plugin
    • Emacs integration
    • JetBrains IDE plugin

  • CI/CD Integration:

    • GitHub Actions workflow
    • GitLab CI integration
    • Pre-commit hooks
    • Automated code review in PRs

  • Collaboration Features:

    • Shared workspaces
    • Team templates and presets
    • Usage analytics dashboard
    • Cost allocation for teams

  • API & SDK:

    • REST API for AIWB operations
    • Python/JavaScript SDK
    • Webhook support
    • Custom provider plugins

  • Multi-Modal Support:

    • Image analysis and generation
    • Diagram generation (Mermaid, PlantUML)
    • Audio transcription for prompts
    • Screen recording analysis

  • Quality Assurance:

    • Security vulnerability scanning
    • Code quality metrics
    • Performance analysis
    • Best practices enforcement

Contributing

We encourage contributors to:

  • Pick up tasks from the roadmap
  • Suggest new features and improvements
  • Report bugs and issues
  • Improve documentation
  • Share templates and workflows

See CONTRIBUTING.md for guidelines.

Feedback

Your feedback is vital to AIWB's success! Please share:

  • Feature requests
  • Use cases
  • Pain points
  • Success stories

Discussions: https://github.com/juanitto-maker/AIworkbench/discussions Issues: https://github.com/juanitto-maker/AIworkbench/issues

This roadmap is a living document and will be updated as the project evolves.