Vigil
Open-source LLM security scanner combining YARA, vectors, and transformers.
About
Vigil is an open-source LLM security scanner that inspects prompts and retrieved context through a multi-stage pipeline: a vector similarity scanner against a jailbreak database, heuristic and YARA-rule scanners, a transformer classifier, and prompt-response similarity analysis, with canary token support.
It ships as both a Python library and a REST API and can be self-hosted, which makes it popular with security researchers experimenting with LLM attack detection. Like all scanner approaches, it is experimental-grade rather than a guarantee of safety.
What it can do
- YARA plus vector plus transformer scanners
- Prompt-response similarity analysis
- Canary token support
- Python library and REST API
Setup
pip install vigil-llm, then configure scanners or deploy the REST API.
Prerequisites
Python, and optionally a vector database for similarity-based detection.
Use cases
Works with
Trust score — 6/10
Scored 6/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 76; Trending score 62/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.