Overview
StarNet is a revolutionary local-first desktop harness that transforms your machine into a vibrant, pixel-art AI agent station. It's not just a simulation; it's a living environment where your AI agents perform real work using actual models and tools. Design your operational space station visually, where rooms represent capability-scoped teams, hallways are authorized handoff lanes, and placed objects grant tangible capabilities. StarNet offers a unique approach to AI agent management, allowing you to scale from a single agent to a concurrent, multi-agent crew, each with its own distinct workspace and permissions.
Key Features
- Living Pixel-Art Station: Visualize your AI agents' operations in a dynamic, pixel-art environment that directly reflects live runtime state.
- Real Work, Real Models: Harness authentic AI models and tools for genuine task execution, not simulated activity. Bring your own API keys (OpenRouter, Anthropic, OpenAI, Google) securely stored in your OS keychain.
- Crew, Not Just Chatbots: Create and manage multiple agents with distinct personas, loadouts, workspaces, transcripts, and bounded permissions for concurrent, independent operations.
- Local-First & Private: Your data, including agent memory, transcripts, and ledgers, persists locally on your machine, ensuring privacy and control.
- Powerful Workflow Design: Visually construct agent workflows by arranging rooms and capabilities. Agents operate based on the layout you define.
- Extensible Capabilities: Integrate with Telegram, Discord, Slack, Signal, Matrix for remote interaction, and attach MCP servers for expanded tool access.
- Tangible Deliverables: Finished work appears in a dedicated OUTBOX as actual files, ready for immediate use.
- Transparent Cost & Performance: Real-time ledgers track spending, budgets, and run history, providing an honest view of AI operations.
- Local Model Support: Run agents using Ollama for a completely free, keyless experience, with all local models being discoverable and usable.
Typical Use Cases
- Automated Task Execution: Set up agents to handle repetitive tasks, from data processing to content generation, operating on schedules or triggered by events.
- Personal AI Assistant Network: Build a team of specialized agents to manage your communications, research, and daily to-do lists, all within a unified, visual interface.
- Creative Content Generation: Design workflows for generating art, writing, or code, leveraging multiple agents working in parallel on different aspects of a project.
- Experimentation and Development: Safely test AI capabilities and workflows locally, with full transparency on model usage and costs, before deploying to production.
- Knowledge Management & Research: Deploy agents to gather, synthesize, and organize information from various sources, creating actionable intelligence.
