benoitgaudin

Exactly. The combination of trace IDs and statement IDs is also a very interesting topic as it has the potential to provide an abstracted view of the runtime behaviours of a system, making is easier to spot when those changes (e.g. as a result of a new version release).

We are investigating how to best leverage coding assistants to inject statement IDs into source code. So far, our experience is that there is no need for MCP tools in order to achieve it. A tool would make sense in order to report the (log statement, statement ID) mapping to backends though and this is something we are planning to add to our MCP server: https://github.com/brontoio/bronto-mcp-server