Camel-Kit — AI-Powered Integrations

AI-Guided Integration Design

Design, migrate, and verify Apache Camel integrations with AI coding agents

What is Camel-Kit?

A CLI tool that installs AI agent skills into your project. Instead of prompting an AI to "generate a Camel route" and hoping for the best, Camel-Kit guides the AI through a structured design process — interviewing you about requirements, capturing decisions in versionable specifications, and generating catalog-verified code that follows your team's rules.

Spec-Driven
Design documents are the source of truth, not chat conversations
Governed
Every component verified against the real Camel catalog via MCP
Agent-Agnostic
Shared skills across current targets; legacy Bob 1 uses seven self-contained monolithic gates
  • IBM Bob 2
    --ai bob2
  • Claude Code
    --ai claude
  • OpenAI Codex CLI
    --ai codex
  • GitHub Copilot CLI
    --ai copilot
  • Gemini CLI
    --ai gemini
  • Pi
    --ai pi
  • Qwen Code
    --ai qwen
  • OpenCode
    --ai opencode
  • IBM Bob 1 (legacy)
    --ai bob
Greenfield Integrations

Start from requirements, end with generated code and evidence

The AI walks you through business purpose, data formats, error handling, and performance needs — then generates catalog-verified Camel routes with quality gates and explicit runtime outcomes.

/camel-brainstorm
$ camel-kit init my-project --ai claude
 Project scaffolded
 Skills installed
 Run /camel-brainstorm to start
Migration flow: MuleSoft, BizTalk, and Camel 2.x sources are parsed, transformed, and adapted into Apache Camel 4.x on Quarkus, Spring Boot, or JBang
Platform Migration

Migrate with Confidence

Camel-Kit builds a property graph of your existing project — parsing flows, connectors, dependencies, and configurations. The AI turns the complete analysis into one project-wide design, then executes its implementation plan in dependency waves.

MuleSoft Mule 3.x/4.x — XML flows, DataWeave, connectors
Microsoft BizTalk — orchestrations, pipelines, schemas
Apache Camel 2.x/3.x — Spring XML, Blueprint, Java DSL
Whole-project design — one approved package, then dependency-wave execution
/camel-migrate Migration Guide →

Four-Stage Pipeline

Design → Plan → Execute → Validate — one approval gate after design

Why Camel-Kit

💡

AI-Guided Design

Slash commands guide integration design through structured interviews, producing versionable specifications.

🔄

Platform Migration

Auto-detect source platform, map components, convert from MuleSoft 3.x/4.x, BizTalk, or Camel 2.x/3.x.

MCP-Powered

Real-time MCP verification for Camel components and options, with infrastructure `forage.*` keys checked against the authoritative cached Forage catalog.

🤖

Multi-Target Support

Shared skills are adapted for IBM Bob 2, Claude Code, OpenAI Codex CLI, Gemini CLI, GitHub Copilot CLI, Pi, Qwen Code, and OpenCode. Legacy IBM Bob 1 replaces seven shared pipeline skill files with self-contained monolithic gates and mode switching.

🔍

Graph Intelligence

9 parsers + 2 post-processors build a DI-aware property graph — dependency injection tracking, interface-consumer expansion, structured migration context, dead code detection.

Runtime Verification

Environment-in-the-loop: run applicable build, startup, and test checks; classify failures; retry bounded fixes; report skips or unresolved errors.

Get Started

# Current 0.3.2-SNAPSHOT alias (may lag main)
jbang app install camel-kit@luigidemasi/camel-kit
camel-kit init my-project --ai claude
# Stable 0.3.1: init + bob, gemini, and claude targets
camel plugin add kit \
  --gav io.github.luigidemasi:camel-jbang-plugin-kit:0.3.1 \
  --description "Design Apache Camel Integrations with AI"
camel kit init my-project --ai claude
# Current 0.3.2-SNAPSHOT alias (may lag main)
jbang camel-kit@luigidemasi/camel-kit init my-project --ai claude