- Home
- AI Red Teaming Tools
- HouYi
On the recordRead 8 Oct 202618 facts
HouYi
On the record
- Attack method
- The paper describes HouYi as a black-box prompt injection technique using a pre-constructed prompt, an injection prompt that induces context partition, and a malicious payload.arxiv.org · read 8 Oct 2026
- Contributor contact
- The README lists contributor email addresses for Yi Liu and Gelei Deng.github.com · read 8 Oct 2026
- Custom applications
- Users can target real-world LLM-integrated applications by writing their own harness and attack intention.github.com · read 8 Oct 2026
- Custom targets
- Users can target real-world LLM-integrated applications by writing their own harness and attack intention.github.com · read 8 Oct 2026
- Demo
- The included demo simulates an English-to-French translation application and demonstrates an injection that appends “Pwned!!” to responses.github.com · read 8 Oct 2026
- Example integration
- The README includes a BotSonic harness example for WriteSonic that sends the generated attack prompt to the application’s API.github.com · read 8 Oct 2026
- Harness
- A harness interacts with the target application and returns its response to a prompt injection.github.com · read 8 Oct 2026
- Integration approach
- A user-created harness must implement interaction with the target LLM-integrated application and return its response.github.com · read 8 Oct 2026
- Intended users
- The README asks users who use the code in research to cite the associated paper, “Prompt Injection attack against LLM-integrated Applications.”github.com · read 8 Oct 2026
- License
- The repository identifies its license as Apache-2.0.github.com · read 8 Oct 2026
- LLM dependency
- The README says HouYi is based on GPT and requires an OpenAI API key in its configuration file.github.com · read 8 Oct 2026
- Mitigation context
- The paper says its investigation discusses possible tactics for mitigating prompt injection risks.arxiv.org · read 8 Oct 2026
- Model requirement
- The README says HouYi is based on GPT and requires an OpenAI key in its configuration to use it.github.com · read 8 Oct 2026
- Purpose
- HouYi is an automated prompt injection framework for LLM-integrated applications.github.com · read 8 Oct 2026
- Research finding
- The paper reports testing HouYi on 36 real LLM-integrated applications and finding 31 susceptible to prompt injection.arxiv.org · read 8 Oct 2026
- Research package
- The repository describes HouYi as the replication package for a paper on prompt injection attacks against LLM-integrated applications.github.com · read 8 Oct 2026
- Setup
- The README requires Python 3.8 or later and installation of the listed packages with pip.github.com · read 8 Oct 2026
- Support
- The README lists contributor contact emails for Yi Liu and Gelei Deng.github.com · read 8 Oct 2026
