shim is an Apache-2.0 command-line tool that detects sensitive data locally, reports on prompts, optionally blocks prompts, masks verified Claude tool events, and watches API traffic. It is intended for developers and security teams managing AI tool usage.
In the AI & LLM security space, shim takes a focused approach. It focuses on preventing sensitive data from leaking through AI prompts, tool events, and API traffic. It is built as an open-source project for developers and security teams using AI coding tools. The project is open source (Apache-2.0). shim is available on the command line, and it can be self-hosted.
GetSHIM builds and maintains shim, and it first shipped in 2026. Development happens publicly on GitHub with 73 commits in the last 90 days. Among its 6 catalogued features are sensitive data detection, prompt reporting, and prompt blocking. It exposes integrations via a public API.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do