https://github.com/confident-ai/deepteam/blob/09f02b5d28926a3169c98b7c75ad93f962f34961/docs/docs/red-teaming-adversarial-attacks-sequential-jailbreaking.mdx?plain=1#L107