15 ICC Framework Prompt Examples (Copy & Paste for ChatGPT, Claude, Gemini)

The ICC framework — Instructions, Context, Constraints — is the fastest way to turn vague prompts into ones that produce exactly what you want. But the best way to learn it isn't theory; it's seeing real examples. This page gives you 15 copy-paste ICC prompts across the tasks people actually use AI for: writing, coding, analysis, email, research, and more. Steal them, swap in your details, and watch your output quality jump.

If you're new to the framework, the short version is this: every strong prompt names what you want done (Instructions), gives the AI the background it needs (Context), and sets the boundaries for the output (Constraints). Our complete ICC framework guide explains the why; this page shows the how. Each example below is structured so you can see the three parts working together.

Writing & Content ICC Prompt Examples

1. Blog post. Instructions: Write a blog post arguing [your position]. Context: I'm [role] at [company type]; the audience is [who]; here's relevant background [details]. Constraints: 800-1000 words, [tone], include [specific elements], avoid [what to avoid].

2. Email newsletter. Instructions: Draft a weekly newsletter section about [topic]. Context: My audience is [who] who care about [what]; our brand voice is [description]. Constraints: Under 200 words, one clear takeaway, end with a single call to action, no jargon.

3. Social post. Instructions: Write a LinkedIn post sharing [insight or story]. Context: I'm a [role] writing for [audience]; the goal is [build authority / start discussion]. Constraints: Under 1,300 characters, conversational, one specific example, end with a genuine question, no hashtag stuffing.

The pattern across all three: the Instructions are specific verbs, the Context gives the AI your situation so it doesn't write generically, and the Constraints prevent the bloated, hedge-everything output AI defaults to. Without context, "write a blog post about remote work" gives you a forgettable essay. With ICC, it gives you something usable.

Coding ICC Prompt Examples

4. Write a function. Instructions: Write a [language] function that [does what]. Context: It will be called from [where]; inputs are [types], expected output is [type]; the codebase uses [conventions]. Constraints: Handle [edge cases], include error handling, add brief comments, follow [style guide].

5. Debug. Instructions: Find and fix the bug in this code [paste]. Context: Expected behavior is [X], actual behavior is [Y]; it fails when [conditions]. Constraints: Explain the root cause before the fix, change as little as possible, don't refactor unrelated code.

6. Code review. Instructions: Review this code for bugs, security issues, and clarity [paste]. Context: This is [what it does], running in [environment]; the team values [readability / performance]. Constraints: Prioritize by severity, give specific line references, suggest fixes don't just flag problems.

Coding prompts especially benefit from ICC because the Context (what the code does, where it runs, the conventions) is exactly what determines whether the AI's output fits your project or needs rewriting. For more on AI coding workflows, see our state of AI coding tools.

Analysis & Research ICC Prompt Examples

7. Data analysis. Instructions: Analyze this data and identify the top 3 insights [paste/describe]. Context: This is [what the data represents]; I need it for [decision/audience]. Constraints: Lead with the most actionable insight, flag any data quality concerns, don't speculate beyond what the data supports.

8. Summarize a document. Instructions: Summarize this document [paste]. Context: I need it to [decide X / brief my team]; I care most about [aspect]. Constraints: Under 200 words, bullet the key points, note anything that requires my attention or action.

9. Compare options. Instructions: Compare [option A] vs [option B] for [my use case]. Context: My priorities are [list]; my constraints are [budget/time/etc]. Constraints: Use a table, give a clear recommendation with reasoning, note where the answer would change based on different priorities.

Email & Communication ICC Prompt Examples

10. Difficult email. Instructions: Draft an email [declining/delivering bad news/pushing back] to [recipient]. Context: The situation is [details]; our relationship is [description]; I want to [preserve relationship / hold firm]. Constraints: Professional but warm, under 150 words, no over-apologizing, propose a path forward.

11. Meeting follow-up. Instructions: Write a follow-up email summarizing this meeting [notes]. Context: Attendees were [who]; the goal was [X]; decisions made were [list]. Constraints: Clear action items with owners, under 200 words, professional tone.

12. Cold outreach. Instructions: Write a cold outreach message to [recipient type]. Context: I offer [what]; their likely pain point is [X]; I found them via [source]. Constraints: Under 100 words, lead with their problem not my pitch, one specific ask, no generic flattery.

Business & Marketing ICC Prompt Examples

13. Marketing copy. Instructions: Write ad copy for [product]. Context: Target audience is [who]; the key benefit is [X]; the brand voice is [description]. Constraints: 3 variations, under 30 words each, lead with the benefit, include a clear CTA.

14. Strategy brainstorm. Instructions: Brainstorm 10 ways to [achieve goal]. Context: My situation is [details]; what I've already tried is [list]; my constraints are [budget/team/time]. Constraints: Prioritize feasibility, skip the obvious, explain the reasoning for the top 3.

15. Product description. Instructions: Write a product description for [product]. Context: It's for [audience] who care about [X]; key features are [list]; it competes with [alternatives]. Constraints: Under 150 words, benefit-led not feature-led, scannable, one strong opening line.

Notice the consistent structure across all 15: the more specific your Context and Constraints, the less the AI has to guess — and guessing is where generic, off-target output comes from. The 30 seconds you spend filling in the brackets saves you the back-and-forth of correcting a vague first attempt.

How to Use These Examples

Pick the example closest to your task, replace the bracketed placeholders with your specifics, and paste it into your AI tool of choice. The framework works identically across ChatGPT, Claude, and Gemini because it's about communicating clearly, not about model-specific tricks. Save the ones you use often as templates — swap the Context each time but keep your Instructions and Constraints, and you build a personal library of reliable prompts.

If you'd rather not structure prompts by hand, the free Prompt Optimizer applies the ICC framework automatically — paste your rough prompt and it returns a structured, optimized version in seconds (3 free per day, no signup). And to use ICC optimization directly inside ChatGPT, Claude, and Gemini without switching tabs, TresPrompt adds one-click optimization to your sidebar.

Frequently Asked Questions

What is the ICC framework?

ICC stands for Instructions, Context, and Constraints — a simple structure for writing effective AI prompts. Instructions state what you want done, Context gives the AI the background it needs, and Constraints set the boundaries for the output (length, tone, format). It works across ChatGPT, Claude, and Gemini. See our full explainer for details.

Do these ICC examples work with all AI models?

Yes. The ICC framework is model-agnostic because it's about communicating clearly, not exploiting model-specific quirks. These examples work identically with ChatGPT, Claude, Gemini, and other major models. Clear instructions, useful context, and explicit constraints help any model produce better output.

How do I turn an ICC prompt into a reusable template?

Keep your Instructions and Constraints fixed, and turn the Context into fill-in-the-blank fields. For example, a blog-post template keeps "write a blog post arguing [position], 800-1000 words, [tone]" and you swap the position and details each time. Save these in a prompt manager so you build a personal library of reliable prompts.

What's the difference between Context and Constraints?

Context is the background information the AI needs to understand your situation — who you are, who the audience is, relevant facts. Constraints are the rules for the output itself — length, tone, format, things to include or avoid. Context shapes what the AI says; Constraints shape how it says it.

Can I automate ICC prompt writing?

Yes. The free Prompt Optimizer applies the ICC framework to your rough prompts automatically — paste what you'd normally type and it returns a structured, optimized version. TresPrompt brings the same optimization directly into ChatGPT, Claude, and Gemini so you can optimize with one click while you type.

Disclosure: Some links in this article are affiliate links. We only recommend tools we've personally tested and use regularly. See our full disclosure policy.