Intent Layer Engineering FAQs

Question 1

How does the skill measure codebase complexity?

Accepted Answer

The skill uses specialized scripts to analyze structure and estimate tokens, identifying semantic boundaries and measuring directory density to suggest where new documentation nodes are needed.

Question 2

When should I create a child AGENTS.md file?

Accepted Answer

You should create a child node when a directory exceeds 20k tokens, when there is a significant shift in responsibility, or when a subsystem has hidden contracts and invariants that need explicit documentation.

Question 3

Can AGENTS.md and CLAUDE.md coexist at the root?

Accepted Answer

No, the skill enforces a single root context file principle. It will prompt you to choose either CLAUDE.md or AGENTS.md as your primary root node to avoid conflicting instructions.

Question 4

Is this skill suitable for existing projects?

Accepted Answer

Yes, it is specifically designed to audit existing repositories and incrementally add the necessary context infrastructure regardless of the project's current documentation state.

Question 5

What is the SME capture protocol?

Accepted Answer

It is a structured interview process used during maintenance mode to extract critical knowledge from Subject Matter Experts, focusing on invariants, scope, and patterns that often confuse new developers.

Intent Layer Engineering

Key Features

Use Cases

Intent Layer Engineering

Key Features

Use Cases