Skip to content

Hypersigil Documentation

Welcome to Hypersigil, a comprehensive testing platform designed to help you develop, refine, and validate AI prompts before deploying them in production environments.

What is Hypersigil?

Hypersigil transforms prompt engineering from an art into a scientific process by providing systematic testing capabilities for AI prompts. Instead of guessing whether your prompts will work effectively, you can test them against various inputs and scenarios, making decisions based on measurable results rather than intuition.

Key Features

🎯 Prompt Development & Testing

  • Create and version your AI prompts with template variable support
  • Test prompts against multiple AI providers (OpenAI, Anthropic Claude, Ollama)
  • Track performance metrics and token usage across different models
  • Multimodal support with file uploads for image analysis and document processing

📊 Comprehensive Test Data Management

  • Organize test cases into logical groups
  • Import test data from various file formats
  • Run batch executions across entire test suites
  • Advanced file management with tagging and search capabilities

🔄 Execution Management

  • Asynchronous processing with real-time status updates
  • Provider-specific concurrency limits and rate limiting
  • Detailed execution history and result analysis
  • Execution bundles for organized batch testing results

💬 Collaborative Features

  • Comment system for execution result analysis
  • Prompt calibration based on feedback and AI-generated suggestions
  • Team collaboration with role-based access control
  • Real-time notifications and event-driven updates

⚙️ Enterprise Ready

  • User management with invitation-based access
  • Database-driven API key management with dynamic provider configuration
  • Comprehensive audit trails and data export
  • Docker deployment with multi-architecture support
  • Onboarding system for guided setup

Getting Started

  1. Introduction - Learn about the platform architecture and workflow
  2. Prompts - Understand prompt creation, versioning, and templating
  3. Executions - Explore the testing and execution system
  4. Test Data - Manage your test datasets effectively
  5. Files - Upload and manage files for multimodal AI testing
  6. Deployments - Create production-ready prompt execution endpoints
  7. Settings - Configure users, API access, and system settings

Quick Start Guide

First Time Setup

When you first access Hypersigil, the system will guide you through creating the initial administrator account. This establishes the foundation for your organization's prompt testing environment.

Basic Workflow

  1. Create Prompts - Define your AI instructions with optional template variables
  2. Prepare Test Data - Import or create test cases organized in groups
  3. Run Executions - Test your prompts against your data using various AI providers
  4. Analyze Results - Review outputs, add comments, and identify improvement opportunities
  5. Iterate & Improve - Use the calibration system to refine your prompts based on results

Architecture Overview

Hypersigil is built with a modern, scalable architecture:

  • Backend: TypeScript/Node.js with Express and SQLite
  • Frontend: Vue.js 3 with TypeScript and Tailwind CSS
  • API: Type-safe REST API with comprehensive validation
  • Providers: Extensible AI provider system supporting multiple services

Support & Contributing

  • Repository: GitHub
  • Issues: Report bugs and request features on GitHub
  • Documentation: This site is built with MkDocs and Material theme

Ready to start testing your AI prompts systematically? Begin with the Introduction to understand the core concepts and workflow.