Name: Harden
Rating: 4.5 (2 reviews)

Harden AIF is a free, local security tool for AI coding agents. Its post-trained model checks tool calls before they run, using your request and session context. It beat frontier models on key agent-security benchmarks, while keeping your repo and tool output on your machine.

It runs entirely local so nothing about my repo or tool output leaves my machine, which is the main reason I'd pick this over a hosted agent-monitoring layer. Setup was a single install and it started flagging risky tool calls immediately without me having to write any rules by hand.

Documentation on how the false-positive threshold is tuned could be clearer, right now it's not obvious what to do when it blocks something you actually meant to run.

CTO of Harden here, thanks Gal, we are going to add better documentation, but currently you can reach the docs at this link http://127.0.0.1:47391/#/docs/config . Currently the docs are bundled inside the binary and they show up as a small link at the bottom right of the settings page http://127.0.0.1:47391/#/settings after you install aif.

@pushpendre_rastogi hey rastogi you shared a localhost link instead of the live link just wanted to give you a heads up

Harden

A security layer for AI coding agents

A security layer for AI coding agents

Harden Reviews

The community submitted 2 reviews to tell us what they like about Harden, what Harden can do better, and more.

What's great

What needs improvement