What is an Intent Layer in the context of this skill?

An Intent Layer is a structured documentation system consisting of a root CLAUDE.md and various AGENTS.md files that define the purpose, rules, and boundaries of a codebase.

What are the prerequisites for using this skill?

The project must have an Intent Layer initialized. If the state is none or partial, you should run the setup skill first.

Can this skill help with large-scale refactoring?

Yes, it can identify cross-cutting concerns and inherited constraints across multiple subsystems, ensuring refactors don't violate global architectural rules.

How does the skill handle missing documentation?

If the Intent Layer lacks an answer, the skill identifies the documentation gap, falls back to code analysis, and flags the response with an 'Uncertain' confidence level.

Does it support parallel searching?

Yes, for large layers with 5 or more nodes, the skill utilizes parallel subagents to search different subsystems simultaneously for faster synthesis.

Intent Layer Query

Name: Intent Layer Query
Author: orban

byorban

•

Learning & Documentation

Queries project architectural documentation to provide instant answers about code ownership, placement, and system constraints.

The Intent Layer Query skill allows developers to treat their codebase documentation as a searchable, intelligent knowledge base. By analyzing a hierarchy of AGENTS.md and CLAUDE.md files, it answers complex architectural questions such as who owns a specific subsystem, where new files should be placed to maintain consistency, and what constraints or pitfalls apply to specific modules. This skill significantly reduces cognitive load during onboarding and refactoring by synthesizing documentation into actionable guidance with explicit confidence levels.

Key Features

01Constraint gathering that merges rules from local nodes up to the global root.

023 GitHub stars

03Placement logic to recommend the correct directory for new features based on project intent.

04Architecture rationale retrieval to explain the 'why' behind design decisions and ADRs.

05Ownership mapping to identify which component is responsible for specific concepts.

06Entry point discovery to find the exact file or function where specific tasks should begin.

Use Cases

01Ensuring new code contributions adhere to existing architectural contracts and patterns.

02Identifying potential pitfalls and anti-patterns before starting a complex implementation.

03Navigating an unfamiliar codebase by querying the hierarchy of responsibilities.

Key Features

01Constraint gathering that merges rules from local nodes up to the global root.

023 GitHub stars

03Placement logic to recommend the correct directory for new features based on project intent.

04Architecture rationale retrieval to explain the 'why' behind design decisions and ADRs.

05Ownership mapping to identify which component is responsible for specific concepts.

06Entry point discovery to find the exact file or function where specific tasks should begin.

Use Cases

01Ensuring new code contributions adhere to existing architectural contracts and patterns.

02Identifying potential pitfalls and anti-patterns before starting a complex implementation.

03Navigating an unfamiliar codebase by querying the hierarchy of responsibilities.