tsfm-anomaly-detect-sensor

id	tsfm_anomaly_detect_sensor
name	TSFM Anomaly Detect Sensor
version	1.0.0
description	Runs integrated TSFM anomaly detection on a provided time-series dataset for explicit target columns.
required_servers	["tsfm"]
asset_types	["chiller","equipment","time_series"]
keywords	["tsfm","anomaly detection","tsad","anomalies","time series"]
default_enabled	true
inputs	{"dataset_path":{"type":"string","required":true},"timestamp_column":{"type":"string","required":true},"target_columns":{"type":"array","required":true},"model_checkpoint":{"type":"string","required":false},"false_alarm":{"type":"number","required":false}}
execution	{"type":"declarative"}

TSFM Anomaly Detect Sensor

Use this skill when the user provides or clearly references an existing time-series dataset and asks to detect anomalies for one or more target sensor columns with TSFM/TSAD.

Do not use this skill when the user asks for forecasting only, when IoT data must first be retrieved, or when the dataset path or target/timestamp columns are missing.

Workflow

1. Run integrated TSFM anomaly detection on the provided dataset and target columns.
2. Use the requested model checkpoint or false-alarm rate when specified.
3. Return the anomaly output file, anomaly count, and status message.

Expected Summary

The final answer should state:

• The dataset path, timestamp column, and target columns used.
• The model checkpoint and false-alarm rate when provided.
• The anomaly count and total record count returned by TSFM.
• The results_file path containing anomaly labels.

Execution Plan

{
  "steps": [
    {
      "name": "anomaly_detection",
      "server": "tsfm",
      "tool": "run_integrated_tsad",
      "arguments": {
        "dataset_path": "$dataset_path",
        "timestamp_column": "$timestamp_column",
        "target_columns": "$target_columns",
        "model_checkpoint": "$model_checkpoint",
        "false_alarm": "$false_alarm"
      }
    }
  ]
}