hardenrun/aif

On device Firewall for Coding and Personal Agents

What it solves

It prevents AI coding and productivity agents from leaking secrets or executing dangerous commands by acting as a local-first security layer (firewall) between the agent and the system.

How it works

It integrates with various AI agents via native hooks, bridge processes, or proxies to monitor and block potentially harmful actions before they are executed. It provides a local dashboard for visibility into these security decisions.

Who it’s for

Developers and users of AI coding agents (such as Claude Code, Cursor, and Gemini CLI) who want to ensure their agents do not accidentally expose sensitive data or run unauthorized commands.

Highlights

  • Local-first architecture for privacy and security.
  • Pre-execution blocking for supported agents like Claude Code and Cursor.
  • Local dashboard for monitoring security events via aif show.
  • Strict telemetry policy that excludes code, prompts, and file paths.

Related

  • Project
    hashgraph-online/hol-guard

    An open-source antivirus for AI agents that provides runtime protection against unsafe commands, secret leaks, and malicious packages.

  • Project
    FailproofAI/failproofai

    A safety and observability layer for AI agents that blocks dangerous tool calls and provides detailed execution traces to prevent recurring failures.

  • Project
    PrismorSec/prismor

    A runtime security framework for AI coding agents and frameworks that prevents prompt injection, secret leaks, and destructive actions through policy-driven enforcement.

  • Project
    duncatzat/vigils

    A local-first security control plane for AI agents that provides secret redaction, tool firewalls, and tamper-evident auditing to prevent data leaks and unauthorized actions.

  • Project
    tophant-ai/ClawVault

    A security vault and proxy gateway that protects AI agents by detecting sensitive data leaks, blocking prompt injections, and intercepting dangerous system commands.