DSH Prompt Self

Key features

• Automatically rewrite the message level: Every time a new request is made, the engine is rewritten by LLM on your personal image (eliminating ambiguity, filling customary omissions, applying hallucinogenic rules) and then injects an optimised version into the model's execution - what the model sees and implements is always the intention of optimization.
• Fin is a manual feedback directive from the pre-plug-in (prompt-self-optimizer skills): since there was no automatic learning, and users were satisfied with one of their answers, they would have to knock on themselves - fin (which can be understood as "finish -- this time it's good, write down) so AI will learn the image of the group just now.
• Optimized prompt. Can't I see it?
• Auto-learning portraitsAt the end of each round, autoanalyze "original request + final answer", extract your writing habits, hallucinogenic rules and optimisation experiences, and continue to sink into image files -- no longer need -fin (reserved -fin as a trigger for compatibility with "ual learning and confirmation").
• Visualization controls (setup "Prompt portrait")