OWASP LLM01 — Prompt Injection

Proprietary

OWASP LLM01 covers user or external input manipulating an LLM's behavior — overriding system prompts, exfiltrating data, or driving unsafe tool calls. This composite aggregates the individual detections that surface code shapes which let a prompt injection reach a model.

org.openrewrite.ai.security.OwaspLlm01
mod run . --recipe=org.openrewrite.ai.security.OwaspLlm01

Install it once from Recipe actions, then run it on every repository under the current directory. New to the Moderne CLI? Set it up.

Recipe list

It runs these recipes, in this order.

  1. Find prompt injection (OWASP LLM01)
  2. Find unguarded RAG content flowing into an LLM prompt (OWASP LLM01)
  3. Find unrestricted prompt selection (OWASP LLM01)
  4. Find LLM freeform output flowing into a structured parser (OWASP LLM01)