The Octopus AI Assistant was released last year to bring a chat based interface into the Octopus web portal. It’s goals were to:
- Provide suggestions on best practices
- Add self-support for failed deployments and runbook runs
- Create sample projects from prompt descriptions
- Provide the ability to answer general questions about your Octopus instance and its configuration
The AI Assistant has been deprecated. The Chrome extension that provided the chat based interface is no longer available for new users, and existing users now see a static message notifying them of the deprecation.
The MCP interface, and specifically the remote MCP interface, is the recommended migration path for users who wish to continue using AI features in Octopus.
For those looking to create projects from prompts, the post End-to-end Kubernetes deployments with the Octopus Deploy MCP server provides example prompts for creating a Kubernetes project in Octopus.
The post Build your own Octopus dashboards with llms.txt and an AI coding agent shows how to use the remote MCP interface to create dashboards and visualizations of your Octopus instance.
The best practices documentation can be consumed by your MCP client to provide suggestions on improving your Octopus instance.
For those looking to include AI features in their deployment processes or runbooks, the Claude Agent Step provides the ability to automate the Claude Code CLI, and also has the ability to interact with the Octopus MCP interface.
Thank you to everyone who has used the AI Assistant and provided feedback - your input has been invaluable in shaping the future of AI features in Octopus.
Happy deployments!
