BFrost
BFrost is a local-first AI operations platform that runs scheduled research, assistant chats, tool-using workflows, approval-gated actions, and worker pipelines entirely on the user's machine.
Ficha actualizada 22 sept 2026, 21:54 · upstream 22 sept 2026, 21:24 · release sin release detectada
Qué es BFrost
Descripción editorial construida a partir del repositorio, su documentación y sus señales de mantenimiento.
BFrost is a self-hosted AI operations platform designed for developers who want full control over their pipeline. Unlike traditional AI tools that require API keys to access models or cloud infrastructure, BFrost runs everything locally — from model inference through to final output. It provides a dashboard for monitoring worker health, managing credentials, and viewing event logs. The core philosophy is simple: every capability is a worker, and the platform only installs, configures, schedules, runs, observes, and uninstalls workers. This means you can add new features by adding new workers without modifying or recompiling the core.
- Schedule daily news digest that researches your topics and posts to Telegram
- Monitor finance news with automated sentiment analysis and market relevance scoring
- Run local AI assistant for personal tasks using LM Studio or Ollama as model provider
- Build project-specific workflows where each step is a separate worker with approval gates
- Create custom scheduled jobs that run specific tools (ffmpeg, search engines) on your machine
- Publish content to multiple platforms through a unified workflow system with version control
- BFrost runs entirely locally — no API keys or cloud subscriptions required for basic functionality
- Every feature is built as a separate worker that can be installed, enabled/disabled, and removed independently
- The platform uses SQLite for persistent state storage with full audit logging of all actions
- Workers are organized into categories (news, research, publishing) making it easy to find relevant tools
- Built-in approval queues allow you to review file changes before they're published or executed
- Supports both local models via LM Studio/Ollama and cloud providers through API key authentication
- The platform requires Node.js 20+ for the core runtime, which may not be available on all systems
- Telegram integration requires a valid bot token that must be kept secret — no built-in encryption
- Voice features depend on external tools like ffmpeg and Whisper models being installed locally
- The dashboard is web-based and requires a browser to access the UI, limiting use in headless environments
- No native mobile or desktop client exists; all interactions must be done through the web interface
Perfil de infraestructura
Estimación inicial derivada del repositorio y de los servicios detectados.
Despliegue detectado
Qué señales reales de instalación hemos encontrado.
Install BFrost by running `npm install` then `npm start`. For production deployment, consider installing it as an OS service using `npm run install-service` so it starts automatically on boot. Use environment variables for sensitive credentials and set up a database backup strategy. The platform is designed to be extensible — new workers can be added without touching the core codebase.
Stack y dependencias
Componentes inferidos a partir de manifests, configuración y documentación.
Puede sustituir a
Relaciones detectadas por el análisis del repositorio y su categoría.
Experiencia real de la comunidad
Ayuda a contrastar la estimación automática con instalaciones reales.
Todavía no hay datos comunitarios para este proyecto.