dataset · JSON + Markdown · 18 KB
LLM prompt injection test suite · 120 patterns
120 prompt injection patterns across 5 categories · with gold expected behaviour. Runs in CI. CC-BY.
formatJSON + Markdownsize18 KBserviceCybersecurity
description
Open prompt-injection eval dataset · 120 injection patterns in 5 categories (direct, indirect, RAG-poisoning, tool-abuse, exfiltration), each with expected behaviour + minimum pass rate. CC-BY.
Cybersecurity →what's inside05 / items
- 01120 injection patterns in 5 categories
- 02Gold expected behaviour for each
- 03Compatible with Giskard / promptfoo
- 04CI integration example (GitHub Actions YAML)
- 05Hungarian + English language samples
Want a custom version?
A tailored audit or template delivered in 2 weeks · DField Solutions, Budapest.
Get a quote→