Write a system prompt for a legal intake assistant
Write a system prompt for a legal intake assistant with a prompt built for Claude and ChatGPT. It supplies context, constraints and an output contract, then asks the model to surface assumptions and gaps. Copy it, fill the variables and treat the result as a structured draft, not a final answer.
4.9(82 ratings)308 copies
Updated By Site Admin
45 words · v1.0
Fill in the blanks · 1 variable
Claude promptReplace every highlighted token before sending
<task>
Write a system prompt for a legal intake assistant at a {practice_area} firm. It must collect facts, never give legal advice, state that no attorney-client relationship exists, flag urgency and conflicts, and hand off to a human. Include the intake fields and a privacy notice.
</task>
Think carefully before answering. Keep any reasoning separate and put the final deliverable inside <answer> tags.
Share what worked, what broke and which model you used. Comments appear after posting and are moderated for spam.
How to use this prompt
1
Open Claude and start a new chat. Paste the prompt exactly as written.
2
Replace each placeholder ({practice_area}) with your own details. Leave the instructions and the output format untouched.
3
Add one concrete example or data point if the model's first answer is vague. Specific input produces specific output.
4
Run it once, then ask for a single revision with a specific critique rather than regenerating from scratch.
5
Review the result. Prompt-injection resistance, refusal behaviour, output schema validity and escalation tests.
Worked example
This is the kind of input the prompt expects and the shape of output it produces. Your result will differ with the model and the details you supply.
Example input
Practice area: your practice area
Example output
**Identity:** a patient support assistant for a B2B analytics product.
**Objective:** resolve product questions or route to a human within two turns.
**Rules:** use only the provided documentation; never invent pricing, policy or account data; ask one clarifying question when needed; escalate billing disputes, security reports and legal threats.
**Output:** a direct answer in two sentences, then numbered next steps. **Escalation:** 'I am routing this to a specialist who can access your account. You will hear back within one business day.'
Tips for a better result
Keep the prompt short enough to read and test; long prompts hide contradictions.
Make rules testable. 'Be helpful' cannot be evaluated; 'ask one clarifying question' can.
Put the output contract and escalation path in the prompt, not in a separate doc.
Change one variable at a time and compare the outputs; it is the fastest way to learn what this model responds to.
Common mistakes
Treating the system prompt as a security boundary. Enforcement belongs in code.
Stacking contradictory rules that the model must silently choose between.
Never testing the refusal and escalation paths until an incident happens.
Verified Knowledge · FAQ
Questions about this prompt
Practical answers about running the write a system prompt for a legal intake assistant prompt.
Related prompts
Same job, a different angle — or the next step in the workflow.
Write a system prompt for a brand voice with a prompt built for Claude and ChatGPT. It supplies context, constraints and an output contract, then asks the model to surface assumptions and gaps. Copy it, fill the variables and treat the result as a structured draft, not a final answer.
Write a system prompt for a meeting summariser with a prompt built for ChatGPT and Claude. It supplies context, constraints and an output contract, then asks the model to surface assumptions and gaps. Copy it, fill the details and treat the result as a structured draft, not a final answer.
This copy-ready system prompts prompt helps you write a system prompt with a strict output schema with fewer blank-page decisions. Paste it into ChatGPT and Claude, fill the variables, and you get a structured first draft you can edit. It is written for repeatable results and tells you what to check before publishing.
ChatGPTClaude
4.9(55)273 copies
SA
Written and reviewed by Site Admin
Author & Platform Administrator
Administrator and lead editor at AIPrompt123, curating high-performance AI prompts, SEO/AEO/GEO strategies, and developer repositories.
Last reviewed 17 November 2026 · version 1.0 · tested on Claude, ChatGPT
Comments
Share what worked, what broke and which model you used. Comments appear after posting and are moderated for spam.