admissions@cyberlawacademy.com | +91-XXXXXXXXXX
DPDPA ResourcesLegal Services
Part 2 of 5

Anatomy of an Effective Prompt

A well-structured prompt is the foundation of successful AI interaction. Learn the essential components that transform vague requests into precise instructions that consistently produce high-quality outputs.

~75 minutes 5 Sections 6 Prompt Templates

2.1 The CRISP Framework

Effective prompts share common structural elements. The CRISP framework provides a memorable checklist for constructing prompts that consistently produce quality outputs: Context, Role, Instructions, Specifications, and Parameters.

The Five Components of an Effective Prompt
C

Context

Background information the AI needs to understand the situation

R

Role

The persona or expertise level you want the AI to adopt

I

Instructions

Clear, specific directives about what you want done

S

Specifications

Details about format, length, style, and structure of output

P

Parameters

Constraints, boundaries, and things to avoid or include

💡 Key Insight

Not every prompt requires all five components. Simple tasks may need only instructions, while complex legal work benefits from the full framework. Learn to identify which components matter most for each task.

2.2 Context: Setting the Stage

Context provides the background information that shapes how the AI interprets your request. Without proper context, even well-written instructions can produce irrelevant results.

What Context Includes

  • Domain: The field or area of expertise (e.g., "Indian contract law")
  • Situation: The specific scenario or problem (e.g., "breach of service agreement")
  • Audience: Who will read the output (e.g., "for a client with no legal background")
  • Purpose: Why you need this (e.g., "to advise on potential remedies")
  • Relevant facts: Key information that affects the analysis
Without Context
"Write about contract breach."

Result: Generic, unfocused content that may not apply to your jurisdiction or situation.
With Context
"Context: A software development company based in Mumbai failed to deliver a project by the agreed deadline. The contract includes a liquidated damages clause of Rs. 50,000 per week of delay. The delay is now 6 weeks."

Result: Targeted analysis specific to Indian law and the actual facts.
⚖️ Legal Practice Tip

When providing context for legal queries, specify: (1) the jurisdiction, (2) the type of parties involved, (3) relevant dates, and (4) any specific contractual or statutory provisions at issue. This grounds the AI's response in applicable law.

2.3 Role: Defining the AI's Perspective

Assigning a role to the AI shapes the tone, depth, and perspective of its responses. The same question answered by a "law professor" versus a "practical litigator" will yield notably different outputs.

Effective Role Assignment

Role Prompting
A technique where you instruct the AI to respond as if it were a specific type of expert, adopting the knowledge, communication style, and perspective associated with that role.

Examples of Useful Roles for Legal Work

  • Senior Corporate Lawyer: For business-focused, practical advice
  • Law Professor: For theoretical analysis and academic depth
  • Opposing Counsel: To identify weaknesses in your position
  • Judge: For balanced analysis of both sides
  • Legal Writer: For clear, accessible explanations
  • Compliance Officer: For regulatory-focused review
Role Assignment Example Effective
Role: Act as a senior corporate lawyer with 15 years of experience in Indian M&A transactions.

Task: Review the following due diligence findings and identify the top 5 legal risks that could affect deal valuation.

Perspective: Advise from the buyer's perspective, focusing on issues that could justify price reduction or walk-away scenarios.
⚠️ Caution

While role assignment improves response quality, remember that AI "playing" a lawyer is not the same as actual legal advice. The AI doesn't have professional duties, malpractice insurance, or accountability. Always apply professional judgment to AI outputs.

2.4 Instructions: Clear Directives

Instructions are the core of your prompt -- what you actually want the AI to do. Vague instructions produce vague results; specific instructions produce specific, useful outputs.

Principles of Effective Instructions

  1. Use action verbs: "Analyze," "Compare," "Draft," "Identify," "Explain" -- not "Tell me about"
  2. Be specific: "List 5 potential defenses" not "What are some defenses?"
  3. Break down complex tasks: Use numbered steps for multi-part requests
  4. Specify what to include AND exclude: "Focus on procedural issues; ignore substantive merits"
  5. Indicate priority: "Most importantly..." or "The key question is..."
Weak Instruction
"Help me with this contract."

Problem: What kind of help? Review? Draft? Explain? The AI must guess.
Strong Instruction
"Review this service agreement and:
1. Identify any provisions that are one-sided against the service provider
2. Flag any ambiguous terms that could lead to disputes
3. Suggest specific language changes for each issue found"

Instruction Templates for Legal Tasks

Contract Review Template
Review the following [contract type] and provide:
1. A summary of key terms (parties, subject matter, duration, value)
2. Identification of unusual or concerning clauses
3. Missing standard protections that should be added
4. Specific redline suggestions for high-risk provisions
5. Overall risk assessment (Low/Medium/High) with justification

2.5 Specifications and Parameters

Specifications control the format and style of output. Parameters set boundaries and constraints. Together, they ensure the AI's response fits your exact needs.

Output Specifications

  • Format: Bullet points, numbered list, prose, table, JSON
  • Length: "In 200 words," "In 3-5 paragraphs," "Brief summary"
  • Structure: Specific sections or headings to include
  • Tone: Formal, conversational, technical, simplified
  • Audience level: Expert, intermediate, layperson

Constraint Parameters

  • What to avoid: "Do not include case citations" or "Avoid legal jargon"
  • Boundaries: "Only consider Indian law" or "Focus on the 2023-2024 period"
  • Requirements: "Must include section references" or "Include practical examples"
  • Exceptions: "Unless specifically relevant to enforceability"
Complete Prompt with Specifications Best Practice
Context: I'm preparing a client memo on force majeure clauses post-COVID.

Role: Act as a senior contracts lawyer familiar with Indian law.

Instructions: Analyze how Indian courts have interpreted force majeure provisions since 2020 and provide practical drafting recommendations.

Specifications:
- Format: Professional memo with headings
- Length: 500-700 words
- Include: Key judicial observations, practical drafting tips
- Tone: Professional but accessible to business clients

Parameters:
- Focus on contract law (not insurance or employment)
- Cite only real cases (note: I will verify all citations)
- Do not include any disclaimer language
✓ Pro Tip

When you need a specific output format, provide an example of what you want. "Format your response like this example: [provide template]" is often more effective than lengthy descriptions.

Key Takeaways

  • The CRISP framework (Context, Role, Instructions, Specifications, Parameters) provides a reliable structure for effective prompts
  • Context grounds the AI's response in your specific situation, jurisdiction, and requirements
  • Role assignment shapes tone, depth, and perspective -- choose roles strategically for your task
  • Clear instructions use action verbs, break down complex tasks, and specify what to include and exclude
  • Specifications control format, length, and style; parameters set boundaries and constraints
  • Not every prompt needs all five components -- match complexity to the task