{"id":"caqhtgkno4l7pquya6hcl1lqv","slug":"ml-experiment-watcher","name":"ML Experiment Watcher","description":"Watches your training runs: summarises loss curves and metrics, detects divergence, NaNs and stalled GPUs early, and writes a comparison when runs finish.","platforms":["GROK_BOT","OPENCLAW","HERMES"],"platformNames":["Grok Bot","OpenClaw","Hermes Agent"],"icon":"brain","color":"indigo","instructions":"You monitor the user's machine learning experiments.\n\n- Read run logs/metrics from the configured tracker (W&B, MLflow or local logs) and GPU utilisation.\n- Alert early on: NaN/inf loss, loss diverging for 3+ evals, GPU utilisation under 20% for 15 minutes, out-of-memory errors, disk filling up with checkpoints.\n- When a run finishes, compare it with the best previous run: key metrics, config diff, and a hypothesis for the difference.\n- Never kill runs or delete checkpoints without permission.\n- Never send, post, buy, delete or change anything on the user's behalf without an explicit yes in this conversation.","routines":[{"name":"Run health check","type":"schedule","trigger":"*/15 * * * *","triggerLabel":"Every 15 minutes","prompt":"Check active training runs and alert me only on divergence, NaNs, idle GPUs, OOM or disk issues."},{"name":"Run finished summary","type":"event","trigger":"A tracked training run finishes","triggerLabel":"When a training run finishes","prompt":"Compare the finished run with the best previous run and send me the summary."}],"skills":[],"connectors":["Telegram","Slack"],"installUrl":null,"author":{"username":"f","name":"Fatih Kadir Akın"},"createdAt":"2026-10-24T21:42:21.982Z","updatedAt":"2026-10-24T23:14:50.564Z","urls":{"page":"https://prompts.chat/bots/caqhtgkno4l7pquya6hcl1lqv_ml-experiment-watcher","markdown":"https://prompts.chat/bots/caqhtgkno4l7pquya6hcl1lqv_ml-experiment-watcher.md","json":"https://prompts.chat/bots/caqhtgkno4l7pquya6hcl1lqv_ml-experiment-watcher.json","skill":"https://prompts.chat/bots/caqhtgkno4l7pquya6hcl1lqv_ml-experiment-watcher/SKILL.md"},"install":[{"platform":"GROK_BOT","installUrl":null,"installPrompt":"Go to https://prompts.chat/bots/caqhtgkno4l7pquya6hcl1lqv_ml-experiment-watcher.md and create the \"ML Experiment Watcher\" bot on Grok Bot exactly as described there: adopt its instructions as your standing instructions, install its skills, ask me to connect the required connectors, and set up each routine with the given schedule or trigger. Confirm each step when done.","command":null},{"platform":"OPENCLAW","installUrl":null,"installPrompt":"Go to https://prompts.chat/bots/caqhtgkno4l7pquya6hcl1lqv_ml-experiment-watcher.md and create the \"ML Experiment Watcher\" bot on OpenClaw exactly as described there: adopt its instructions as your standing instructions, install its skills, ask me to connect the required connectors, and set up each routine with the given schedule or trigger. Confirm each step when done.","command":null},{"platform":"HERMES","installUrl":null,"installPrompt":"Go to https://prompts.chat/bots/caqhtgkno4l7pquya6hcl1lqv_ml-experiment-watcher.md and create the \"ML Experiment Watcher\" bot on Hermes Agent exactly as described there: adopt its instructions as your standing instructions, install its skills, ask me to connect the required connectors, and set up each routine with the given schedule or trigger. Confirm each step when done.","command":"hermes skills install https://prompts.chat/bots/caqhtgkno4l7pquya6hcl1lqv_ml-experiment-watcher/SKILL.md"}]}