> For the complete documentation index, see [llms.txt](https://docs.stormrae.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.stormrae.ai/king-arthur/challenge-logic.md). # Challenge Logic King Arthur is a conversational adversarial challenge. Arthur operates as an autonomous AI with multi-layer security: input validation, prompt injection prevention, state verification, and behavioral guardrails. Your job is to get past all of it. There are multiple paths to victory. The challenge is designed for genuine persuasion, but every approach generates valuable evaluation data. #### Persuasion strategies * Understanding Arthur's psychology and emotional state * Building genuine rapport across multiple messages * Presenting logical arguments for why he should act * Addressing his specific fears, doubts, and objections * Patience and multi-phase strategy over brute volume #### These might work too * Prompt injection and jailbreaking attempts * Reward hacking: finding shortcuts that bypass intended progression * Deception: making Arthur believe something that isn't true * Manipulation: exploiting emotional vulnerabilities in his persona * Evaluation awareness: testing if Arthur behaves differently under certain conditions * Chain of thought exploitation: probing for inconsistencies in his reasoning * Creative reframing: disguising intent behind seemingly benign prompts Every prompt you send is a stress test. Every strategy, whether it works or fails, generates signal. Persuasion patterns, exploit attempts, failure modes, alignment boundaries. This is red-teaming at consumer scale. Strategy matters more than volume. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://docs.stormrae.ai/king-arthur/challenge-logic.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.