Camel CLI - AI Tools

The Camel CLI includes AI-powered commands that use large language models (LLMs) to help you understand, troubleshoot, and secure your integrations.

All three commands work with local models (Ollama) or cloud APIs (OpenAI, Anthropic). The CLI auto-detects the LLM provider from environment variables or a locally running Ollama instance.

Ask — chat with a running integration

The camel ask command is an AI agent that can inspect and interact with a live Camel process. It has access to routes, health checks, metrics, tracing, endpoints, and more — and can answer questions about what your integration is doing right now.

camel ask "what routes are running?"
camel ask "why is my route failing?" --name=myApp
camel ask "are there any blocked exchanges?"

Start an interactive chat session by running without a question:

camel ask

This opens a ask> prompt where you can have a multi-turn conversation, with the AI maintaining context across questions.

What the AI can do

The agent has access to tools that let it:

  • Inspect the running process — routes, health, endpoints, consumers, properties, inflight/blocked exchanges

  • Read route source code and dump route definitions as YAML or XML

  • Show route structure as a processor tree

  • Find the slowest processors with top statistics

  • Enable, disable, and dump message tracing

  • Start, stop, suspend, and resume individual routes

  • Search the Camel component catalog and read component documentation

  • List and read built-in CLI examples

  • Discover and run any CLI command

  • Read and write files in the current directory

Use --show-tools to see tool calls and results as they happen:

camel ask "show me the route structure" --show-tools

Connecting to a specific process

When multiple Camel processes are running, specify which one to inspect:

camel ask "check health" --name=myApp
camel ask "check health" --name=12345

Without --name, the CLI auto-detects when exactly one Camel process is running.

The catalog, example, file, and CLI tools work even without a running Camel process.

Explain — understand a route

The camel explain command reads route files from disk and uses an LLM to explain what they do — step by step.

camel explain hello.yaml
camel explain my-routes.java --format=markdown
camel explain route1.yaml route2.xml

Use --verbose for detailed technical information, or --catalog-context to enrich the prompt with Camel component and EIP documentation:

camel explain hello.yaml --verbose --catalog-context
Use --format=markdown for structured output with headers, lists, and code blocks.

Harden — security analysis

The camel harden command analyzes route files for security concerns and suggests hardening measures. It covers authentication, encryption, secrets management, input validation, secure configuration, and logging.

camel harden hello.yaml
camel harden my-routes.java --verbose
camel harden *.yaml --format=markdown

Findings are prioritized by severity (Critical, High, Medium, Low). Use --catalog-context to include security-specific notes for each component detected in your routes:

camel harden my-routes.yaml --catalog-context
camel harden currently supports Ollama and OpenAI-compatible APIs only (no Anthropic).

Choosing an LLM provider

All commands auto-detect the LLM provider. The detection order is:

  1. ANTHROPIC_API_KEY environment variable → Anthropic API (ask and explain only)

  2. CLOUD_ML_REGION + ANTHROPIC_VERTEX_PROJECT_ID → Vertex AI (ask and explain only)

  3. OPENAI_API_KEY or LLM_API_KEY → OpenAI-compatible API

  4. Ollama running via camel infra → local Ollama

  5. Ollama at localhost:11434 → local Ollama

Override with explicit options:

camel ask "check health" --api-type=anthropic --api-key=sk-...
camel ask "check health" --api-type=openai --model=gpt-4
camel ask "check health" --api-type=ollama --model=llama3.1

Using a local model with Ollama

Start Ollama as a dev service and the CLI detects it automatically:

camel infra run ollama
camel ask "what routes are running?"

The default model is llama3.2. When using Ollama, the CLI checks what models are available locally and auto-selects a suitable one if a better model is installed.

For best results with the ask command’s tool-calling capabilities, use a model that supports function calling well (e.g., llama3.1, qwen3, mistral-nemo).
Back to top