LLM Server Live Log 📜

Live log for the Local LLM Server: every request the chat client makes, with its status, how long it took, the tokens it cost and how full the context now is - plus loads, unloads, errors and the parameter repairs on the way through. One row per request, opened when it arrives and closed when it finishes, so a cold start looks like a load rather than a hang. Drop it anywhere on the canvas - it needs no connections, and it fills itself in from the backlog, so a chat that happened while ComfyUI was closed is still there. Message text is NOT logged unless the server node's 'log_text' is on.