SILO Framework
Welcome to the documentation for SILO — the trust layer for modern AI tool execution.
What is SILO?
SILO (Secure, Isolated, Lightweight, Offloaded) is a robust execution engine and package manager for AI Agent tools (skills). It solves the "last mile" problem of agentic workflows by providing a secure, predictable environment for tools to run.
Core Pillars
- 🔒 Isolation: Skills are run in an encapsulated manner, meaning each skill contains all its dependencies and holds them in memory only for the duration of its execution. This requires no manual
venvinitialization or dependency installation, as it happens automatically. Deleting a skill clears all cached dependencies. The core idea is encapsulating all its data into a single "container". - 🛡️ Security: Secrets are encrypted at rest and injected via STDIN. They never leak to the LLM or logs.
- 🧩 Discovery: Agents find tools semantically using BM25. No more hallucinating tool names.
- ✨ Premium DX: Built-in browser-based approvals and interactive auth flows.
Quick Start
Install the core framework using uv:
Initialize your first skill:
Why SILO?
SILO helps developers bring their ideas to life by conquering the complexity of AI tool development. It provides a trusted starting point for every agent integration, ensuring that tools are discoverable, secure, and reproducible across any environment.