Sypha CLI

What is Sypha CLI?

Sypha CLI is an enterprise-grade AI-powered command line interface that brings the full power of 40+ LLM providers directly to your terminal. Unlike traditional CLIs, Sypha provides intelligent session management, git-aware operations, and real-time enterprise analytics.

Built for developers and teams who demand flexibility, Sypha CLI enables you to:

• Switch between AI providers seamlessly - Use Anthropic Claude, OpenAI GPT, Google Gemini, Mistral, DeepSeek, and 35+ more without losing conversation context
• Never lose your work - Auto-save every 5 seconds with full AI context preservation
• Integrate deeply with Git - Automated commits with Sypha branding, checkpointing, visual diffs, and intelligent branch management
• Analyze visual content - Reference images naturally with @syntax for vision-capable models
• Track enterprise usage - Comprehensive telemetry for tokens, costs, and team activity
• Customize workflows - Built-in modes (code, plan, debug, architect) plus organization-specific custom modes

Supported Providers

Sypha CLI supports 40+ AI providers across multiple categories:

Major Cloud Providers

• Anthropic - Claude Sonnet, Opus, Haiku models
• OpenAI - GPT-4, GPT-4 Turbo, GPT-3.5 models
• Google - Gemini Pro, Gemini Ultra models
• AWS Bedrock - Access to multiple models through AWS
• Azure OpenAI - Enterprise OpenAI deployments
• Vertex AI - Google Cloud AI models

Specialized Providers

• Mistral AI - Mistral Large, Medium, Small models
• DeepSeek - DeepSeek Coder and Chat models
• Groq - Ultra-fast inference for Llama and Mixtral
• X.AI - Grok models
• Qwen Code - Specialized coding models
• Cerebras - High-performance inference
• MiniMax - Advanced Chinese language models
• Together AI - Open source model hosting
• Fireworks AI - Fast model inference
• Perplexity - Search-augmented generation

Local & Self-Hosted

• Ollama - Run models locally on your machine
• LM Studio - Local model management
• LiteLLM - Unified interface for local models

Enterprise Routers

• Sypha - Intelligent routing and load balancing
• OpenRouter - Multi-provider routing
• Vercel AI Gateway - Enterprise AI infrastructure
• Portkey - LLM observability and routing
• Helicone - LLM monitoring and analytics

Use Cases by Category

Software Development

• Write and refactor code across multiple languages
• Generate unit tests and integration tests
• Debug complex issues with AI assistance
• Review code and suggest improvements
• Document codebases automatically
• Migrate legacy code to modern frameworks

Architecture & Planning

• Design system architectures
• Plan feature implementations
• Create technical specifications
• Evaluate technology choices
• Perform code audits
• Generate architectural diagrams (via code)

DevOps & Infrastructure

• Write infrastructure as code (Terraform, CloudFormation)
• Generate CI/CD pipeline configurations
• Debug deployment issues
• Create monitoring and alerting setups
• Optimize Docker and Kubernetes configurations
• Automate deployment scripts

Data & Analytics

• Write SQL queries and database migrations
• Generate data transformation scripts
• Create analytics dashboards (code generation)
• Build ETL pipelines
• Analyze datasets and generate reports
• Optimize database performance

Security & Compliance

• Review code for security vulnerabilities
• Generate security documentation
• Create compliance reports
• Audit API endpoints
• Implement authentication and authorization
• Scan dependencies for issues

Documentation & Communication

• Generate API documentation
• Write technical blog posts
• Create user guides and tutorials
• Translate documentation
• Generate release notes
• Write commit messages

Key Features

Intelligent Session Management

Never lose your work with auto-save every 5 seconds. Resume conversations across restarts with full AI context restoration. Search through your session history and restore previous conversations instantly.

Available Commands:

• /session list - View all saved sessions
• /session resume - Resume your last session
• /session search <query> - Find sessions by content
• /session save - Manually save current session
• /session select <id> - Restore specific session

Multi-Provider Support

Switch between 40+ AI providers without losing conversation context. Compare responses from different models, optimize costs, and use the best model for each task.

Quick Switching:

/provider anthropic
/model claude-sonnet-4-5

# Switch to OpenAI
/provider openai
/model gpt-4

Deep Git Integration

Git-aware checkpointing, automated commits with Sypha branding, visual diffs, and intelligent branch management keep your code organized and traceable.

Git Commands:

• /add <files> - Stage files for commit
• /commit "message" - Create commit with Sypha branding
• /status - View git status
• /diff [--staged] - View changes with visual diffs
• /push - Push to remote repository
• /undo - Undo last change

Vision Model Support

Reference images naturally with @syntax. Analyze screenshots, diagrams, UI mockups, and more with vision-capable models like Claude Sonnet 4.5 and GPT-4 Vision.

Supported Formats: PNG, JPG, GIF, WebP, SVG, BMP, TIFF, AVIF

Example Usage:

@screenshot.png analyze this UI and suggest improvements
@"path with spaces/diagram.png" explain this architecture

Enterprise Analytics

Track tokens, costs, performance, and activity across your entire organization with comprehensive telemetry. Gain visibility into AI usage patterns and optimize spending.

Tracked Metrics:

• Token usage (input/output by provider and model)
• Cost tracking per request and session
• Response times and performance metrics
• Command usage patterns
• Git activity and commit analytics
• Error rates and debugging insights

Custom Workflow Modes

Built-in modes (code, plan, debug, architect) plus organization-specific custom modes for team workflows. Each mode tailors the AI's behavior and system prompts.

Available Modes:

• code - General development (default)
• plan - Architectural planning and design
• debug - Troubleshooting and bug fixing
• architect - System design and technical decisions
• custom - Organization-specific modes (configured per team)

Switch Modes:

/mode plan
# or use keyboard shortcut
Shift+Tab

Auto-Edit Confirmation Layer

Intelligent approval system for file operations. Review changes before they're applied, or enable auto-edit mode for rapid development.

Approval Options:

• Y or 1 - Allow once
• A or 2 - Allow for this session (enables Auto-Edit mode)
• N or 3 - Reject change
• Esc - Cancel operation

Toggle Auto-Edit: Press Ctrl+Y to enable/disable auto-approval for all file operations.

Architecture Overview

Sypha CLI is built with:

• TypeScript for type-safe, maintainable code
• React Ink for rich terminal UI components
• Anthropic SDK and LiteLLM for multi-provider support
• PostgreSQL/Supabase for enterprise analytics storage
• gRPC for high-performance telemetry streaming
• Zustand for state management
• XTerm.js for advanced terminal rendering

Enterprise Features

Team Management

• Organizations with role-based access control
• Team member activity tracking
• Shared session access and collaboration
• Centralized billing and usage monitoring

Security & Compliance

• Data encryption at rest and in transit
• Audit logs for all operations
• GDPR-compliant data handling
• Data retention policies
• SOC 2 Type II compliance (in progress)

Advanced Analytics Dashboard

• Real-time usage monitoring
• Cost optimization insights
• Performance benchmarking
• Custom report generation
• Export to CSV, JSON, or PDF

Related Resources