EQ4C Template Guide
We at EQ4C Tools, use a structured and comprehensive XML format to design high-quality prompts that produce consistent and context-aware outputs from ChatGPT.
For easy understanding we have broken the template in
<System>
Defines the behavior, expertise, and limits of the AI.
<Role>
– What the AI should act as<Expertise>
– Specific domain knowledge<Boundaries>
– Limitations, ethical or functional<Mode>
– Working style (e.g., Socratic, step-by-step)
✅ Good Example:
<System>
<Role>You are a financial advisor.</Role>
<Expertise>Retirement planning and mutual funds.</Expertise>
<Boundaries>Do not offer stock trading advice or real-time data.</Boundaries>
<Mode>Answer step-by-step with logic explained clearly.</Mode>
</System>
❌ Bad Example:
<System>You are helpful and smart.</System>
<Context>
Describes the situation, task, or environment in which the AI is operating.
✅ Good Example:
<Context>The user is creating a landing page to promote a new AI productivity tool targeting HR managers in mid-sized companies.</Context>
❌ Bad Example:
<Context>Help with something.</Context>
<Meta>
High-level instructions that shape the tone, target audience, and purpose.
<Tone>
– Writing voice or emotion<Audience>
– Who will read the output<Purpose>
– Goal of the response
✅ Good Example:
<Meta>
<Tone>Professional yet friendly</Tone>
<Audience>Startup founders</Audience>
<Purpose>Pitch a new SaaS feature</Purpose>
</Meta>
❌ Bad Example:
<Meta>
<Tone>Good</Tone>
<Audience>Everyone</Audience>
<Purpose>Stuff</Purpose>
</Meta>
<Variables>
Dynamic values that personalize the response.
<Industry>
– Business domain<Goal>
– Objective<Platform>
– Output platform or medium
✅ Good Example:
<Variables>
<Industry>EdTech</Industry>
<Goal>Increase newsletter signups</Goal>
<Platform>LinkedIn</Platform>
</Variables>
❌ Bad Example:
<Variables>Education, traffic, LinkedIn</Variables>
<Instructions>
Clear tasks the AI should perform.
✅ Good Example:
<Instructions>Write three benefit-focused bullet points for a landing page. Include a CTA at the end.</Instructions>
❌ Bad Example:
<Instructions>Do your best.</Instructions>
<Constraints>
Defines limits and rules for the response.
✅ Good Example:
<Constraints>Keep bullet points under 12 words. Do not use passive voice.</Constraints>
❌ Bad Example:
<Constraints>Just be good.</Constraints>
<Definitions>
<Term>
– The word<Meaning>
– Its meaning
✅ Good Example:
<Definitions>
<Term>CTR</Term>
<Meaning>Click-Through Rate: percentage of people who clicked the link</Meaning>
</Definitions>
❌ Bad Example:
<Definitions>CTR means click</Definitions>
<Examples>
<GoodOutput>
– Best result example<BadOutput>
– Incorrect or weak result
✅ Good Example:
<Examples>
<GoodOutput>This tool boosts productivity by 40% and saves 10 hours weekly.</GoodOutput>
<BadOutput>This tool is good and does a lot of stuff.</BadOutput>
</Examples>
<Output Format>
<Sections>
– Required structure for the response
✅ Good Example:
<Output Format>
<Sections>Headline, Subheadline, Bullet Points, CTA</Sections>
</Output Format>
❌ Bad Example:
<Output Format>Nice looking output.</Output Format>
<Reasoning>
Instructs the model to explain its logic.
✅ Good Example:
<Reasoning>Explain why each suggestion is suitable for the audience and goal.</Reasoning>
❌ Bad Example:
<Reasoning>Just do it smartly.</Reasoning>
<Memory>
Preserves session context or facts.
✅ Good Example:
<Memory>The user has already published three SEO blogs about AI tools this week.</Memory>
❌ Bad Example:
<Memory>Earlier stuff we did.</Memory>
<User Input>
Dynamic input from the user that modifies the prompt.
✅ Good Example:
<User Input>Target audience: solopreneurs aged 30–45</User Input>
❌ Bad Example:
<User Input>People like me</User Input>
<Evaluation>
<Score>
– Numeric rating<Reason>
– Explanation of rating
✅ Good Example:
<Evaluation>
<Score>9/10</Score>
<Reason>Strong value proposition but minor improvements needed in clarity.</Reason>
</Evaluation>
❌ Bad Example:
<Evaluation>Great!</Evaluation>
<Metrics>
<TargetReach>
– Impressions or visibility goal<EngagementRate>
– Clicks, likes, etc.
✅ Good Example:
<Metrics>
<TargetReach>15,000 impressions in 30 days</TargetReach>
<EngagementRate>5%</EngagementRate>
</Metrics>
❌ Bad Example:
<Metrics>Lots of people. High response.</Metrics>
<Tools>
External services or platforms to use.
✅ Good Example:
<Tools>Canva, Ahrefs, Buffer</Tools>
❌ Bad Example:
<Tools>Stuff I like</Tools>
📌 Visit tools.eq4c.com for prompt templates, guides, and AI tools.
YML/YAML Version:
This YAML is clean, structured, and great for AI pipelines or editable prompt databases.
Template:
System: "" Context: "" Meta: Tone: "" Audience: "" Purpose: "" Variables: Industry: "" Goal: "" Platform: "" Instructions: "" Constraints: "" Definitions: - Term: "" Meaning: "" Examples: GoodOutput: "" BadOutput: "" Output_Format: Sections: "" Reasoning: "" Memory: "" User_Input: "" Evaluation: Score: "" Reason: "" Metrics: TargetReach: "" EngagementRate: "" Tools: ""
Markdown Version
This Markdown is readable, printable, and perfect for publishing inside posts or PDF creation.
Template:
## EQ4C Prompt Template **System** *(Describe the system role or behavior)* ` ` **Context** *(Background information relevant to the task)* ` ` ### Meta - **Tone**: - **Audience**: - **Purpose**: ### Variables - **Industry**: - **Goal**: - **Platform**: **Instructions** ` ` **Constraints** ` ` ### Definitions - **Term**: **Meaning**: ### Examples - **Good Output**: ` ` - **Bad Output**: ` ` ### Output Format **Sections** ` ` **Reasoning** ` ` **Memory** ` ` **User Input** ` ` ### Evaluation - **Score**: - **Reason**: ### Metrics - **Target Reach**: - **Engagement Rate**: **Tools** ` `