monorepo-navigation

Enforce scoped navigation of monorepo codebases using Monorepo Map and area AGENTS.md boundaries.

Updated Apr 11, 2026
One-click install
npx skills add https://github.com/bennybennison/agent-toolkit --skill monorepo-navigation-bennybennison
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: monorepo-navigation
Source: https://github.com/bennybennison/agent-toolkit/tree/main/skills/monorepo-navigation
Command: npx skills add https://github.com/bennybennison/agent-toolkit --skill monorepo-navigation-bennybennison

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Prevents agents from randomly exploring or modifying unrelated parts of a monorepo by providing a clear routing and boundary discipline so automated agents act within intended areas and respect architectural layers.

Core Features & Use Cases

  • Monorepo Map: A routing table in the repository AGENTS.md that maps areas to paths and their AGENTS.md files so agents know where to look.
  • Area Boundaries: Per-area Imports / Exports / Boundaries sections that document allowed dependencies and explicit "Never Touches" to avoid architectural shortcuts.
  • Cross-Area Change Protocol: A stepwise workflow that identifies affected areas, starts changes from the innermost layer outward, respects import direction, and updates AGENTS.md as public surfaces change.
  • Use Case: Safely add a new domain entity and propagate required repository adapters, API models, and frontend changes while preventing connectors from importing API internals.

Quick Start

Focus on the src/connectors area, read its AGENTS.md and Monorepo Map, and only implement changes that respect the declared imports, exports, and boundary rules.

Frequently Asked Questions about monorepo-navigation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I focus AI agents on specific areas of a monorepo codebase?

You can focus agents on specific monorepo areas by using a Monorepo Map and per-area AGENTS.md files that declare boundaries, imports, and exports to prevent exploration of unrelated paths.

How do I prevent automated agents from modifying unrelated files in a large repository?

To prevent agents from modifying unrelated files, enforce scoped navigation using a routing table in the repository AGENTS.md and explicit "Never Touches" boundary declarations in each area's AGENTS.md file.

What is the best way to manage cross-area changes in a monorepo without breaking architectural layers?

The best way to manage cross-area changes is a stepwise protocol that identifies affected areas, starts from the innermost layer outward, respects import directions, and updates AGENTS.md as public surfaces change.

When do I need to use a Monorepo Map for agent navigation?

You need a Monorepo Map when working in multi-area repositories to provide agents with a routing table that maps areas to paths and their respective AGENTS.md files for safe, scoped task execution.

Can I use AGENTS.md files to enforce import and export boundaries between different repository areas?

Yes, per-area AGENTS.md files document allowed dependencies and explicit boundaries, preventing architectural shortcuts like connectors importing API internals during refactors or feature implementation.

Does monorepo navigation work for tasks like adding new domain entities and updating API models?

Yes, monorepo navigation supports safely adding domain entities and propagating required repository adapters, API models, and frontend changes while respecting declared area boundaries and import rules.