multiagent-iot-to-tsfm-forecast

id	multiagent_iot_to_tsfm_forecast
name	Multiagent IoT To TSFM Forecast
version	1.0.0
description	Retrieves IoT history for one explicit asset and runs TSFM forecasting on an explicit dataset path.
required_servers	["iot","tsfm"]
asset_types	["chiller","equipment","time_series"]
keywords	["iot","tsfm","forecast","multiagent","chiller"]
default_enabled	true
inputs	{"site_name":{"type":"string","required":true},"asset_id":{"type":"string","required":true},"start":{"type":"string","required":true},"final":{"type":"string","required":true},"dataset_path":{"type":"string","required":true},"timestamp_column":{"type":"string","required":true},"target_columns":{"type":"array","required":true},"model_checkpoint":{"type":"string","required":false},"forecast_horizon":{"type":"integer","required":false}}
execution	{"type":"declarative"}

Multiagent IoT To TSFM Forecast

Use this skill when the user asks for the common two-agent workflow: retrieve IoT history for one asset and forecast one or more sensor columns with TSFM. The user must provide or clearly imply a dataset path that TSFM can read, because the current IoT history tool returns observations but does not write a TSFM-ready CSV file.

Do not use this skill for anomaly detection, work-order reasoning, failure-mode ranking, or prompts that do not provide enough IoT and TSFM arguments.

Workflow

1. Retrieve IoT history for the requested asset and time window.
2. Run TSFM forecasting on the provided dataset path and requested target columns.
3. Return both the IoT retrieval trace and the TSFM forecast output path.

Expected Summary

The final answer should state:

• The IoT asset, site, and time window used.
• The number of IoT observations returned.
• The TSFM dataset path, timestamp column, target columns, and model checkpoint.
• The forecast results_file path returned by TSFM.

Execution Plan

{
  "steps": [
    {
      "name": "iot_history",
      "server": "iot",
      "tool": "history",
      "arguments": {
        "site_name": "$site_name",
        "asset_id": "$asset_id",
        "start": "$start",
        "final": "$final"
      }
    },
    {
      "name": "forecast",
      "server": "tsfm",
      "tool": "run_tsfm_forecasting",
      "arguments": {
        "dataset_path": "$dataset_path",
        "timestamp_column": "$timestamp_column",
        "target_columns": "$target_columns",
        "model_checkpoint": "$model_checkpoint",
        "forecast_horizon": "$forecast_horizon"
      }
    }
  ]
}