Hi Camel Community, As the way we build integrations evolves, we’ve been looking closely at how Apache Camel can better support the next generation of tooling—specifically, how developers and AI agents can seamlessly collaborate.
We are seeing a shift toward prompt-first development, where visual tools act as a display layer, and AI agents assist with writing, modifying, and debugging routes. However, there’s a gap: Humans and AI agents currently lack a shared, structured view of a running Camel application. While you might look at Karavan, Kaoto, or a dev console, an AI agent is stuck parsing raw logs or looking at file system changes. To bridge this gap and enable blazing-fast feedback loops, we’ve opened a new proposal: https://issues.apache.org/jira/browse/CAMEL-23613 What We Are Proposing: - A Shared Runtime Contract: A unified API/protocol exposing route topology, endpoint inventory, metrics, and error states that both humans and AI tools can understand. - Standardizing, Not Reinventing: Evaluating existing standards like the Model Context Protocol (MCP) and seeing how we can supercharge the existing Camel LSP. - Local-First, Fast Feedback: Enabling a workflow where you prompt an agent, Camel hot-reloads the change, and both you and the AI can instantly verify the health of the application on your laptop. We Need Your Input! We don't want to design this in a vacuum. We want to know how you build platforms, products, or custom tooling on top of Camel. * How are you currently embedding or extending Camel in your own platforms? * Are you already experimenting with AI agents, LLMs, or MCP tools to generate or manage Camel routes? * What does your ideal "human-plus-AI" developer workflow look like? Please take a look at the ticket, leave your thoughts, use cases, or critiques. Let’s build the future of integration together! Best regards, -- Otavio R. Piske http://orpiske.net
