Repository navigation
📢 [Project Sharing] GNS3 Copilot: AI/Natural Language Automation for Network Topologies! #2583
Closed
yueguobin
started this conversation in
Show and tell
Replies: 2 comments 8 replies
|
Hello, is it possible to use a local LLM through something like Ollama/LM Studio? |
1 reply
|
Update: GNS3 Copilot has been merged into the GNS3 3.0 branch and will be part of the upcoming 3.1.0 release. Thanks everyone for the interest and support! |
7 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Hello GNS3 Community! 👋
I'm excited to share a project I've been developing called GNS3 Copilot. My goal is to integrate AI capabilities into GNS3 to make setting up, troubleshooting, and learning network experiments more intuitive and efficient.
✨ What GNS3 Copilot Does
Automation Backbone: It utilizes popular automation tools like Netmiko/Nornir/Telnet to execute device commands.
Topology Operations: I've extended the gns3fy module by modifying key parts and encapsulating tools to perform topology operations (reading topologies, creating devices, and configuring links).
GNS Server 3.0 Support: Crucially, I've also modified the gns3fy module to support GNS server 3.0 (the changes are currently housed in the test directory).
Natural Language UI: Initially, I used it for knowledge review and quickly setting up experiments. I later built a simple UI using Streamlit, which allows users to manipulate topologies, troubleshoot, and learn network concepts using natural language commands.
❓ Seeking Community Feedback
I believe this AI-powered functionality could be a valuable candidate for integration directly into the official GNS3 Web UI.
I have already sent a formal integration proposal to the core development team (at developers@gns3.net), but I'm sharing this here to gather wider community input.
What do you think of controlling your GNS3 labs using natural language?
What other AI-assisted features would you like to see?
I look forward to hearing your thoughts and feedback!
Repository Address: https://github.com/yueguobin/gns3-copilot
All reactions