Intent-Driven Engineering Works With Any AI Tool (Not Just Claude)
- Mark Kendall
- Mar 7
- 4 min read
Intent-Driven Engineering Works With Any AI Tool (Not Just Claude)
Introduction
One of the most common questions I received recently from consultants around the world was surprisingly practical:
“Mark, this is great — but what if our client won’t allow Claude Code?”
This question makes perfect sense. In a large consulting organization, teams work across hundreds of clients, industries, and countries. Every company has its own approved toolchain. Some use GitHub Copilot. Some allow Cursor. Some only allow internal AI systems.
So the real question becomes:
Is Intent-Driven Engineering tied to a specific tool — or is it a universal way of working?
The answer is important.
Because once you understand the principle, you realize something powerful:
Intent-Driven Engineering works with any AI tool.
Tools may change, but the method remains the same.
What Is Intent-Driven Engineering?
Intent-Driven Engineering is a development workflow where the work begins with clearly defined intent before any code is written.
Instead of jumping directly into implementation, engineers first establish:
the goal
the constraints
the architectural context
the definition of success
Only after that intent is clearly defined does implementation begin.
The process typically follows a simple structure:
Intent → Context → Plan → Execute → Verify
This changes the role of AI tools from code generators into engineering collaborators.
The Old Model: Tool-Driven Development
For decades, software development has followed a fairly chaotic pattern.
A developer gets an idea, opens an IDE, and starts writing code. Problems are solved as they appear. Systems evolve organically.
The workflow often looks like this:
Idea
↓
Open IDE
↓
Write Code
↓
Fix Errors
↓
Refactor
↓
Repeat
Over time, this creates familiar problems:
tightly coupled systems
architectural drift
inconsistent patterns
“spaghetti” integrations between services
One system calls another system, which calls another system, which calls another system — and eventually the architecture becomes difficult to reason about.
Many teams simply try to manage this complexity through experience and documentation.
But AI introduces a new possibility.
The New Model: Intent-Driven Development
In an intent-driven workflow, engineers slow down briefly before coding begins.
The first step is establishing the intent of the change.
For example:
Goal:
Create an API endpoint to retrieve customer orders.
Constraints:
Must follow company API standards.
Must support pagination.
Must integrate with OrderService.
Success Criteria:
Endpoint returns orders within 200ms.
This intent acts as a north star for the work.
Once intent is defined, the engineer provides context:
existing architecture
related services
coding standards
dependencies
With intent and context available, an AI assistant can then help generate a structured implementation plan.
Instead of asking:
“Write the feature.”
The workflow becomes:
Step 1 — Create controller
Step 2 — Implement service layer
Step 3 — Add validation
Step 4 — Write tests
Step 5 — Integrate with existing systems
Each step is executed iteratively.
Finally, the system verifies whether the original intent has been satisfied.
Why Claude Code Feels So Powerful
Tools like
were designed around the idea of understanding developer intent.
This is why many engineers experience a dramatic productivity boost when they first try it.
However, the key insight is this:
Claude Code did not invent the idea of structured engineering workflows.
It simply automates them extremely well.
The Method Works With Any Tool
Intent-Driven Engineering can be used with virtually any modern AI development tool.
Examples include:
The difference is simply how much the tool automates the process.
Some tools help more with planning.
Some help more with code generation.
Some provide repository awareness.
But the core workflow remains identical.
The Portable Process (Works Anywhere)
Consultants can implement Intent-Driven Engineering with a simple structure inside any repository.
1. Define Intent
Create a file that defines the purpose of the work.
Example:
This file describes:
the goal
constraints
architecture expectations
success criteria
2. Capture Context
Next define the system environment.
Example:
This might include:
architecture diagrams
service relationships
coding standards
dependencies
This gives the AI assistant the environment it needs to reason effectively.
3. Generate a Plan
Ask the AI system to create a structured plan based on the intent and context.
For example:
Based on the intent and context,
create a step-by-step implementation plan.
4. Execute Incrementally
Work through the plan step by step rather than generating everything at once.
This creates:
smaller changes
easier testing
better architectural alignment
5. Verify the Intent
Finally, confirm whether the implementation satisfies the original intent.
Ask questions such as:
Did we achieve the goal?
Are performance constraints met?
Does the solution match the architecture?
This final step closes the loop.
Why This Matters for Consultants
For consultants working across multiple clients, the ability to use a tool-agnostic engineering method is extremely valuable.
Clients may restrict which AI tools are allowed.
But they rarely restrict how engineers structure their thinking.
Intent-Driven Engineering provides a portable workflow that can be applied in any environment.
This means consultants can:
improve development discipline
accelerate delivery
maintain architectural consistency
leverage AI tools responsibly
Regardless of which specific tools are approved.
Intent-Driven Engineering and the Future of Software Development
The real shift happening in software engineering is not simply the introduction of AI coding assistants.
The deeper shift is moving from:
Tools → Code
to
Intent → Systems → Code
In this new world, engineers spend less time writing boilerplate code and more time defining clear architectural intent.
Ironically, this makes architectural thinking more important than ever.
The best engineers will not simply be those who write the most code.
They will be those who define the clearest intent.
Key Takeaways
• Intent-Driven Engineering is a method, not a specific tool.
• Tools like Claude Code, Copilot, and Cursor are simply different implementations of AI-assisted development.
• The core workflow is portable:
Intent → Context → Plan → Execute → Verify
• Consultants can apply this process in any client environment, regardless of tool restrictions.
• As AI becomes more powerful, clear engineering intent will become the most valuable skill in software development.
Comments