Claude Code Plugin

The noodlbox plugin for Claude Code provides skills, commands, and agents that enable intelligent codebase exploration and analysis.

Installation

Automatic Setup (Recommended)

The easiest way to install the plugin is through the noodlbox CLI:

noodl configure

This automatically:

Installs the noodlbox Claude Code plugin
Configures MCP server connection

If the plugin is already installed, noodl configure --force will update it to the latest version.

Manual Setup

If you prefer manual installation:

claude plugin marketplace add noodlbox/noodlbox-claude-plugins
claude plugin install noodlbox@noodlbox

The plugin auto-updates when you run noodl update.

Plugin Architecture

The plugin provides three types of content:

Type	Purpose	Execution
Skills	Knowledge Claude uses for decision-making	Internalized by Claude
Commands	Slash commands you invoke	Triggered with `/noodlbox:command`
Agents	Isolated exploration tasks	Spawned by commands for heavy work

Commands

Plugin commands use the /noodlbox:command format. These are the same as MCP prompts but with Claude Code's plugin namespacing:

Plugin Command	MCP Equivalent
`/noodlbox:init`	`/mcp__noodlbox__init`
`/noodlbox:generate_map`	`/mcp__noodlbox__generate_map`
`/noodlbox:detect_impact`	`/mcp__noodlbox__detect_impact`

`/noodlbox:init`

Experimental - This feature is still in development.

Initialize human-readable labels for communities and processes.

/noodlbox:init [repository]

Creates .noodlbox/labels.json with descriptive names for:

Communities - "Authentication System" instead of "community_42"
Processes - "User Login Flow" instead of "handleLogin_to_validateToken"

The labels file is used by the CLI, MCP resources, and architecture generation.

`/noodlbox:generate_map`

Generate comprehensive architecture documentation.

/noodlbox:generate_map [repository]

Creates an ARCHITECTURE/ directory with:

README.md - Overview, stats, mermaid diagram, module descriptions
{process-slug}.md - Individual files for each key process

Output structure:

ARCHITECTURE/
├── README.md              # Main overview + diagram
├── user-login-flow.md     # Process documentation
├── checkout-process.md
└── ...

If .noodlbox/labels.json exists, uses human-readable names for modules and processes.

`/noodlbox:detect_impact`

Analyze git changes and understand their ripple effects.

/noodlbox:detect_impact [change_scope]

Change scopes:

Scope	Git Equivalent	Use Case
`unstaged`	`git diff`	Before staging
`staged`	`git diff --staged`	Before commit
`all`	`git diff HEAD`	All uncommitted (default)
`compare`	`git diff <base>`	Compare with branch

Output includes:

Changed symbols with file locations
Impacted processes and their execution flows
Risk level based on centrality scores
Cross-community impact analysis
Recommendations for testing

Agents

Agents handle heavy exploration in isolated context to avoid polluting your conversation.

Codebase Analyst

Adaptive exploration agent that scales based on codebase size:

Scale	Communities	Strategy
Small	< 10	Full exploration
Medium	10-50	Strategic sampling
Large	> 50	High-level overview

Used by /noodlbox:generate_map and /noodlbox:detect_impact for large codebases.

Labeller

Generates human-readable labels for communities and processes. Used by /noodlbox:init (experimental).

Skills

Skills teach Claude when and how to use Noodlbox tools effectively. They're automatically loaded with the plugin.

Skill	When Claude Uses It
Exploring Codebases	"How does X work?", "What's the architecture?"
Tracing Call Graphs	"What calls X?", "Trace the flow from A to B"
Assessing Change Impact	"What will this break?", "Check the blast radius"
Generating Documentation	"Document the architecture", "Create a codebase map"

Claude automatically selects the right skill based on your question - no need to invoke them manually.

CLI Depth Control

The CLI supports depth control for query and search commands:

# Get symbol definitions only (default)
noodl query "authentication flow" --depth definitions

# Get symbols with all relationships
noodl query "error handling" --depth full
noodl search "payments" -d full

Depth	Description
`definitions` (default)	Symbol signatures and full body content
`full`	Symbols plus all relationships (calls, imports, edges)

Workflow: First-Time Setup

For a new repository:

1. noodl analyze /path/to/repo    # Analyze codebase
2. /noodlbox:generate_map          # Generate architecture docs

Workflow: Pre-Commit Review

Before committing changes:

1. /noodlbox:detect_impact staged  # See what's affected
2. Review impacted processes
3. Plan testing based on impact

Labels File

The .noodlbox/labels.json file stores human-readable names:

{
  "version": "1.0",
  "repository": "owner/my-app",
  "communities": {
    "abc123": {
      "label": "Authentication System",
      "description": "Handles login, token validation, sessions"
    }
  },
  "processes": {
    "proc_001": {
      "label": "User Login Flow",
      "description": "Validates credentials, creates session"
    }
  }
}

Used by:

noodl CLI commands
MCP resources (@noodlbox:map://)
Architecture generation

Updating the Plugin

Automatic Update (Recommended)

The plugin is automatically updated when you update the CLI:

noodl update

Or force update via configure:

noodl configure --force

Manual Update

claude plugin marketplace update noodlbox
claude plugin update noodlbox@noodlbox

Getting Started - Installation and MCP setup
Change Analysis - Dependencies and impact workflow
Code Exploration - Exploration patterns

On this page