Replies: 1 comment
|
Thank you for sharing the lineage experiment, especially for distinguishing CBM's graph search from your additional correlation logic. A small public example showing the source repositories, query, expected lineage, and observed result would be useful evidence. We have not validated the reported token reduction or cross-repository accuracy, but the concrete workflow is valuable feedback. |
0 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.
Uh oh!
There was an error while loading. Please reload this page.
I have been building a personal multi-agent data lineage discovery project using CBM-MCP graph search to generate focused context for each agent.
CBM-MCP, combined with additional preprocessing and correlation logic, has helped reduce the amount of source code passed to the agents and lower token usage.
The results have been promising so far with public sample repositories, including cross-repository data flows. I am continuing to validate and improve the solution.
I will replicate the same for GUI based ETL tool such as informatica , optim, validate and provide feedback.
All reactions