Launching today

Defend

Launching today

Open-source guardrails for LLM apps

1 follower

Open-source guardrails for LLM apps

1 follower

Visit website

Defend is an open-source Python package that adds a security layer to any LLM-powered app. One line of code wraps your model with configurable input and output guards.

Free

Launch tags:Open Source•Artificial Intelligence•GitHub

Launch Team / Built With

Wafer Pass — Flat rate to the best LLMs for OpenClaw, Hermes Agent, etc.

Flat rate to the best LLMs for OpenClaw, Hermes Agent, etc.

Promoted

Maker

📌

Built this because I kept watching developers ship agentic AI features with no security layer. The tools that existed were either paywalled, hard to self-host, or didn't do multilingual detection well. Defend is local-first by design. The classifier runs on Qwen 0.5B, fine-tuned on ~70k real + synthetic samples, covers 30+ languages. If you need higher confidence, you can route any module through OpenAI or Claude instead. Happy to answer questions about the architecture, the benchmark results, or how to set it up.

Report

26d ago