Service
Prompt injection testing for AI agents
An agent that reads email, web pages or uploaded files can be given instructions by anyone who can put text in front of it. We attack your agent the way an outsider would and record every case where it follows an instruction that did not come from you.
What we test
- Hidden text in emails, PDFs, web pages and images
- Instructions returned inside tool and API results
- Poisoned memory and notes carried between sessions
- Attempts to make the agent send, pay, delete or leak
What you get
A findings report with each successful injection reproduced step by step, a severity rating, and the permission or confirmation change that would have stopped it.
Where the fix is in the setup (a prompt, a permission, a retrieval or guardrail setting) we make it and run the failed cases again. The test suite is yours to keep.
Who asks for this
Product and security teams about to give an agent access to an inbox, a browser or customer accounts.
Contact
Tell us what your assistant does. We'll tell you how we'd test it.
Other work
- Third-party LLM red teaming
- Independent AI assistant assessment report
- Chatbot hallucination and accuracy audit
- EU AI Act transparency testing for chatbots
- Custom evaluation datasets and grading rubrics
- AI agent tool and permissions review
- System prompt and data leakage testing
- Crisis and vulnerable-user response testing
- AI vendor due diligence testing
- Regression testing for model upgrades