OpenC3 steps into the age of AI with our latest release - COSMOS 7.2 Enterprise! Welcome to a faster, smarter, more beautiful COSMOS enhanced by a brand new MCP server, COSMOS sidekick in our new AI chat, and a Grafana overhaul resulting in stunning dashboards and faster performance.

Here's what to look forward to when upgrading to COSMOS 7.2:

BYOAI

COSMOS 7.2 embraces the power of AI by adding a full-featured MCP server. Instead of injecting an OpenC3-specific LLM, COSMOS 7.2 is BYO-AI. Point you favorite LLM to /mcp to quickly and easily access the COSMOS APIs. Once

"COSMOS, Build Me a Dashboard"

COSMOS 7.2 also features a brand new, built-in sidebar that lets you access your favorite LLM right from inside COSMOS. Now you can have Claude build you a dashboard, ChatGPT interpret a data table, or Gemini whip up a graph.

‍

Security Note

Using the chat will send data to the LLM (an external server if you connect to a public LLM like Claude, or OpenAI). Until configured, no data will be sent to any LLM. You can also use local OpenAI compatible LLMs like Gemma, if run by llama-server.

‍

Go Full Throttle with Grafana

Speaking of graphs, with COSMOS 7.2 we completely overhauled Grafana. Now you can build shiny new dashboards faster and easier, pushing Grafana to the max with COSMOS.

‍

Additional COSMOS 7.2 Enhancements

• Websockets updated to not pass token as query parameter and instead as part of subscribe
• Operator is now throttled to starting 5 new microservices every 5 seconds (by default). This prevents a huge CPU/RAM/IO burst at initial startup.
• Script messages updated to use a Redis Stream instead of pub/sub. This allows replay of script messages on connect to a running script and ensures messages are never lost to the frontend on initial start

COSMOS 7.2 Release Notes

• COSMOS Enterprise Release Notes
• COSMOS Core Release Notes

‍

Stay connected with OpenC3 and make sure you always get the latest COSMOS updates by subscribing to our newsletter!