Has Octave developed an MCP server/connector for BricsCAD yet?

You don't need an MCP server for such things. BricsCAD plugins can do almost anything you want, and Claude is perfectly capable of writing plugins.

You may want to take a look at the JTB World plugins, BatchAttEdit and SmartBatch may be the plugins worth looking at. BatchAttEdit runs internally in BricsCAD and SmartBatch is used externally (i.e. outside BricsCAD) and its scripts call on BricsCAD to do the changes/updates.

Like the others, write your own lisp to update the layout names the way you want. I have a couple, as an example reorder the display of the layouts so now update layout names to match.

For us job number was dwg name, sheet number you can get using VL code returns layout number, revision look at title block attributes and find last revision, eg PA PB A B 1 2 and so on.

Something to get you started.

; rename layouts by AlanH

(defun c:renlayouts ( / laylst lays lay)
(setq laylst '())

(setq dname (cadr (fnsplitl (getvar 'dwgname ))))
(setq lays (vla-get-layouts (vla-get-activedocument (vlax-get-acad-object))))
(vlax-for lay lays
(setq laylst (cons (list (vla-get-taborder lay) (vla-get-name lay)) laylst))
)

(foreach lay laylst
(if (= (car lay) 0)
(princ)
(command "layout" "R" (cadr lay) (strcat dname "D-" (rtos (car lay) 2 0)))
)
)
(princ)
)
(c:renlayouts)

@gizmon, that's awesome! I think it might be cool to build some sort of harness, or tools the model can call in process

Here's a proof of concept

import lmstudio as lms
from pyrx import Ap, Db, Ge, Ed
import json

# missing in BRX, maybe LOGFILE
def tool_get_command_history():
    print("tool_get_command_history_called")
    return Ed.Core.getLastCommandLines(32, True)

tool_get_command_history.__doc__ = (
    "Retrieves the last lines of the CAD command line log to inspect errors or past commands."
)

def tool_evaluate_lisp(statement: str):
    print("tool_evaluate_lisp_called {}".format(statement))
    lock = Ap.AutoDocLock()
    result = Ed.Core.evaluateLisp(statement)
    return f"tool_evaluate_lisp_result: {result}"

tool_evaluate_lisp.__doc__ = (
    "Evaluates an AutoLISP statement within the active CAD document context. "
    "Accepts a 'statement' string parameter containing the LISP code. "
    "Example: \"(entmake (list '(0 . \"LINE\") '(8 . \"0\") '(100 . \"AcDbEntity\") "
    "'(100 . \"AcDbLine\") (cons 10 '(0.0 0.0 0.0)) (cons 11 '(10.0 0.0 0.0))))\""
)

@Ap.Command(Ap.CmdFlags.SESSION)
def doit():
    ps, user_prompt = Ed.Editor.getString(1, "\nWhat should the agent do? ")
    if ps != Ed.PromptStatus.eNormal or not user_prompt:
        return

    model = lms.llm("qwen/qwen3.8-27b")
    print("\nAgent is deciding which CAD tool to use...")
    try:
        response = model.act(
            user_prompt, tools=[tool_get_command_history, tool_evaluate_lisp], on_message=print
        )
        print(f"\nFinal Agent Summary: {str(response)}")
    except Exception as ex:
        print(f"\nAn error occurred during SDK communication: {str(ex)}")

Result

It draws the line!

other tools could be, tool_pyload, tool_pyreload, tool_appload a tool for Ed.Core.cmdS(). The model could generate Lisp, Python routines to load a run.