20 of the best prompts for stop Claude from giving generic responses, step by step across 4 stages. Works with ChatGPT, Claude, and Gemini.
20 of the best prompts for stop Claude from giving generic responses, step by step across 4 stages. Works with ChatGPT, Claude, and Gemini.
Get Claude to produce specific, original, genuinely useful answers instead of the safe, surface-level responses it defaults to. This guide walks you through every stage of Stop Claude From Giving Generic Responses, from Understand Why It Keeps Being Generic all the way through Lock In Specificity for Ongoing Conversations, with a curated, copy-ready prompt at each step. Each stage targets a specific phase of the process so you always know exactly what to ask and what output to expect. Works with ChatGPT, Claude, and Gemini and any other major AI tool.
Generic responses are not random. Claude defaults to them for predictable reasons. These prompts help you diagnose what is triggering the generic mode so you can shut it down at the source.
You just gave
You just gave me a very generic response. I want you to analyze it: which parts are actually specific to my situation, and which parts would you give to any person asking this question? Then tell me what additional context you would need to give me a genuinely useful answer.
Rate each paragraph
Here is your response: [PASTE RESPONSE]. Rate each paragraph on a scale of 1 to 5 for how specific it is to my situation versus how generic it is. Explain your ratings and tell me how to get you to a 5 across the board.
After each version
I am going to ask you the same question three times with increasing specificity. After each version, explain how your answer would change. This will help me understand what information makes the biggest difference. Question: [YOUR QUESTION]
Before you answer
Before you answer, list all the assumptions you are making about me, my experience level, my situation, and my goals. I will correct the wrong ones so your actual answer is not built on false premises.
Specific information about my
What specific information about my situation would most change your answer? Ask me the three questions that would have the biggest impact on the quality of your response.
The most reliable way to get specific responses from Claude is to load your prompt with so much context that there is no room for generalities. These prompts show you how.
Person is me:
I need you to answer this as if advising a specific person, not the average person. That person is me: [DESCRIBE YOUR ROLE, SITUATION, CONSTRAINTS, WHAT YOU HAVE TRIED, WHAT MATTERS MOST]. Now answer with that specific person in mind: [YOUR QUESTION]
Before answering
Before answering, I want to give you context that makes my situation unusual. [DESCRIBE WHAT IS DIFFERENT: YOUR BACKGROUND, THE SPECIFIC PROBLEM, PRIOR ATTEMPTS, CONSTRAINTS, GOAL]. With this in mind, what would you do in my position?
Answer could only apply
Give me an answer that could only apply to someone in my situation. My situation: [DESCRIBE WITH SPECIFICS]. Not general principles. Not what works for most people. What works for me given exactly these constraints.
Most concrete answer
I want the most concrete answer you can give. I do not want suggestions, principles, or frameworks. I want the first three specific actions I should take tomorrow morning. My situation: [DESCRIBE]
Already expert
I am already an expert in [YOUR FIELD] at the level of [DESCRIBE YOUR EXPERIENCE]. Calibrate your response accordingly. Skip the basics and give me the insight that would actually be new information at my level.
Claude is trained to be careful and nuanced, which can produce wishy-washy responses when you need a clear recommendation. These prompts push it past the hedging.
Do not need
I do not need a balanced assessment. I need you to give me a clear recommendation for my situation. Take a position and defend it. My situation: [DESCRIBE]. My question: [YOUR QUESTION]
If you had
If you had to advise someone exactly like me and you could only give one piece of guidance, what would it be? Not a list. Not considerations. The single most important thing. My situation: [DESCRIBE]
Tell something
Tell me something about this topic that most people in my situation get wrong, that is counterintuitive, or that you would normally hold back because it is not the safe answer. My situation: [DESCRIBE]
Position you think
I want the perspective of someone who has strong opinions on this topic, not a neutral summary. Take the position you think is most defensible and make the case for it directly. Topic: [DESCRIBE]
What would someone
What would someone with deep expertise in [FIELD] see in my situation that a generalist would miss? I want the specialist read, not the general one. My situation: [DESCRIBE]
Over a long conversation, Claude can drift back to generic responses. These prompts keep it anchored to your specific context throughout.
Before we start
Before we start this conversation, I want to establish the context that should shape everything you tell me. I am [ROLE AND SITUATION]. My constraints are [LIST]. I already know [WHAT YOU KNOW]. What matters most to me is [PRIORITIES]. Now, with that grounding, my first question: [YOUR QUESTION]
You just gave
You just gave me a generic answer. Please rewrite it using only the specific details of my situation that I gave you. Everything in your response should be something that only applies to someone with my exact profile.
Flag anything
Rule for this conversation: every recommendation must be something I can execute within [TIMEFRAME] given these constraints: [LIST CONSTRAINTS]. Flag anything that falls outside those constraints instead of suggesting it.
Situation has specific
My situation has specific characteristics that make standard advice wrong: [DESCRIBE WHAT IS DIFFERENT]. Every answer you give should account for these. If standard advice would not apply, say so and give me the version that does.
Build situation
Build a profile of my situation based on everything I have told you. Then use it to rewrite your last response so it is genuinely personalized to someone with exactly my profile.
Claude prioritizes safety and broad usefulness by default, which can produce even-handed responses when you want specificity. The key is not just giving context but framing the prompt so Claude understands you want a specific recommendation, not a general overview. Telling it explicitly to take a position and skip the basics helps significantly.
Tell it directly: do not give me a balanced view, give me your recommendation. You can also add: if you had to advise someone exactly like me, what would you tell them? Claude respects explicit framing instructions and will shift from balanced to directive when asked clearly.
Claude does not know your expertise level unless you tell it. Adding something like: I have been doing this for ten years, skip the fundamentals and go straight to [specific topic] signals Claude to calibrate its response to your actual level.
More specific context almost always helps, but the type of context matters most. The most useful context is: your specific constraints, what you have already tried, what you already know, and what outcome you actually need. Generic context like "I am a busy professional" does less than specific context like "I have 20 minutes and no budget."
Establish a context profile at the start of the conversation and refer back to it when Claude drifts. You can also ask Claude to check its own response: which parts of this answer are specific to my situation versus generic? That forces it to catch and correct its own drift.
AI Prompts for Deep Analysis Request for Claude
Claude sometimes provides surface-level answers instead of the in-depth analysis requested.
See promptsAI Prompts for Diagnosing and Correcting Fact Errors
Claude sometimes confidently presents incorrect information, including invented citations and wrong dates, which can mislead users.
See promptsAI Prompts for Stop Claude From Sounding Like an AI
Get Claude to write in a natural, human voice that passes the "did a person write this?" test every time..
See prompts