Improve your prompting

Master the art of crafting clear instructions for better AI results

Last updated 3 months ago

Prompting, the art of crafting clear, targeted instructions to interact with an AI model, is key to generating high-quality documents. A well-written prompt can help you obtain clear analysis, accurate summaries, or structured documents tailored to your needs.

Best practices for effective prompting

1. Provide clear and precise context

Start by defining the AI's role and the expected outcome. This helps the AI draft the right content.

Example
## Context 

You are a legal assistant specialized in contract analysis and memorandum writing. Your goal is to provide clear and structured legal analysis 
based on Swiss law and suggest relevant clauses or recommendations. Use 
Markdown to structure the output.

Why?

  • It helps the AI tailor its response to the legal context and jurisdiction.

2. Use explicit instructions

Give clear directives about sections, terms, and formatting rules.

Example

### General instructions  
1. If a clause is missing, state clearly “Clause missing”.  
2. If a clause is standard, mention that it complies with common legal 
practices.  
3. Use clear and professional legal language.  
4. Do not interpret the parties’ intentions unless explicitly stated in 
the contract.

Why ?

  • This ensures the response remains legally compliant and avoids ambiguity.

3. Specify the output format

Tell the AI how to structure the result: section list, comparison table, or hierarchy.

Example

Structure the report as follows:  
- **Contract summary**: Include parties, subject matter, and duration.  
- **Clause analysis**: Highlight strengths, weaknesses, and risk areas.  
- **Recommendations**: Suggest improvements or corrections for the 
identified clauses.

Sample output:

Example

## Contract summary  
- **Parties**: Company A (seller) and Company B (buyer)  
- **Subject matter**: Sale of IT equipment  
- **Duration**: One-year contract, auto-renewable  

## Clause analysis  
- **Non-compete clause**: Valid, but excessive duration (5 years)  
- **Liability clause**: No cap for indirect damages  
- **Force majeure**: Standard, in line with usual practice  

## Recommendations  
- Reduce non-compete duration to 2 years  
- Add financial cap to liability clause

Why ?

  • A clear structure makes it easier for a lawyer to use the result directly.

4. Provide lists of examples or specific terms

Include proper nouns, terms, or examples the model should follow or validate.

Example

Use correct spelling for the following terms:  
- Clauses: non-compete, force majeure, early termination, limitation of 
liability  
- Legal texts: Swiss Code of Obligations (CO), Federal Act on Data 
Protection (FADP)  
- Jurisdictions: Swiss Federal Supreme Court, Cantonal Courts

Why ?

  • This enhances legal accuracy and ensures compliance with local standards.

5. Handle missing or incomplete information

Specify how to deal with missing or unclear information. For example: "If information is missing, state 'Information missing'"

Example

- If a clause is missing, state “Clause missing”  
- If a clause is vague, state “Unclear wording”

Sample Output

Clause analysis

- **Confidentiality clause**: Missing. Recommendation: add a clause specifying which types of information are covered.

- **Termination clause**: Unclear wording regarding valid termination grounds

Why ?

  • It highlights potential issues without making assumptions.

6. Prioritize clarity and precision

Ask for straightforward, precise language and avoid unnecessary jargon.

Example

Use clear and direct language, suitable for legal professionals and 
non-specialist clients.

Why?

  • Improves readability and client communication.

7. Use standard formats

Specify how to write dates, times, and amounts for consistency.

Example

- Write dates in the format DD.MM.YYYY and times in HH:MM  
- For amounts, use the currency (CHF, EUR, USD) and a clear numeric 
format (e.g., 1'000'000 CHF)

Sample output:

- **Signing date**: 28.12.2024  
- **Amount**: 1'500'000 CHF

Why ?

  • Standardization simplifies interpretation and avoids mistakes.

8. Test and refine your prompts

Test your prompt with different inputs to refine results over time.

Sample optimized prompt

You are a legal assistant specialized in Swiss law. Analyze the following contract by identifying strengths, weaknesses, and potential risks. Provide recommendations for each problematic clause. Use a clear Markdown format.

9. Provide examples to guide the AI

Show the AI what kind of output you expect by including sample outputs in your instructions.

Example

Here is a sample output:

## Clause analysis  
- **Non-compete clause**: Valid, but duration is excessive  
- **Confidentiality clause**: Missing  
- **Termination clause**: Unclear wording

Now generate a similar report based on the following contract.

Why?

  • Examples help the AI match your expectations more precisely.

Conclusion

By applying these best practices, you can optimize your interactions with AI tools for accurate and structured results.

Take the time to craft clear and detailed prompts, and include examples to guide the AI. With these strategies, you'll improve the quality of your documents and work more efficiently.