steve02081504/fount

What it solves

fount is a programmable AI agent runtime platform that allows users to create, customize, and deploy AI characters with deep logic control. It bridges the gap between high-efficiency productivity tools (like coding assistants) and immersive emotional interaction (like role-play characters), providing a standardized environment to avoid the "reinventing the wheel" problem when building agents.

How it works

The platform uses a modular architecture based on Deno and Express, where different components are treated as "parts" that can be installed and customized. Key modules include:

• Characters (chars): The core personality and logic of the agent.
• Worlds: Knowledge bases that influence decisions and chat history.
• ** uma Personas:** User profiles that can modify perceptions and words for immersive roleplay.
• Shells: Integration layers that extend agents into IDEs, browsers, Telegram, Discord, and terminals.
• AI Source Generators: JavaScript-based templates that allow the platform to connect to virtually any AI provider (OpenAI, Claude, OpenRouter, etc.) via custom API routing.

Who it’s for

• Developers and Geeks: Those who want to customize agent logic via code rather than just prompts, and who need features like real-time code execution and Git-driven management.
• AI Character Creators: Users seeking high-fidelity, immersive interactions with support for rich text, images, and group chats.
• Professional Users: People needing a private, P2P encrypted hub for team collaboration and report generation.

Highlights

• Real-time Code Execution: Live compilation and running of multiple languages (C/C++, Rust, Python, JS) directly within chat blocks.
• P2P Collaboration: Native encrypted communication over LAN, Bluetooth, or custom protocols for private team interaction.
• Deep Integration: Ability to embed characters into JetBrains, Neovim, Zed, browser extensions, and terminal shells.
• AI-Assisted Creation: Tools to generate full personas from a single sentence and a CI/CD pipeline (fount-charCI) for character development.
• Universal AI Connectivity: Custom JS-based generators to connect to any AI source, including native API pool support for intelligent routing.