Reuse a subprocess-per-stage Python dashboard for a second pipeline via module overrides
To run a second pipeline on an existing stdlib http.server job dashboard without copying it, expose a few module-level hooks (STAGES, default target, port, run-record filename, HTML path, and the CHILD_SCRIPT path used for --child-stage subprocesses), then in a new entry file import dashboard as dash and reassign dash.STAGES/dash.run_stage/dash.video_detail etc. before calling dash.main(). The non-obvious part: if each stage runs in a child process launched as python <file> --child-stage, the child must be the NEW entry file (set CHILD_SCRIPT to its __file__), or the subprocess runs the original pipeline's run_stage. Also make the front end render stages by a view field sent from the server rather than hard-coded stage keys, so one HTML page serves both pipelines.