菜单
Detects anomalies in time series data using TimeGPT. Identifies outliers, level shifts, and trend breaks without model training. Use when identifying anomalies, outliers, or unusual patterns in time series. Trigger with "detect anomalies", "find outliers", "anomaly detection".
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/majiayu000/claude-skill-registry --skill nixtla-anomaly-detector命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
LLM token logprobs and calibration. Per-decision confidence, ECE, Brier, reliability diagrams, low-confidence triage.
Analyze LLM token logprobs and calibration. Use for per-decision confidence, ECE, Brier scores, reliability diagrams, and low-confidence triage.
回顾最近 N 天的 Claude Code 使用记录——扫描原始会话数据，按主题分组汇总"我都做了什么"，并从个人操作系统视角输出模式、风险与增删建议。当用户说 /recap、"看看我这几天做了什么"、"回顾一下我最近的会话"、"这两天我用 claude 干了啥"、"活动回顾" 时使用。
基于 SOC 职业分类
正在显示 SKILL.md
| name | nixtla-anomaly-detector |
| description | Detects anomalies in time series data using TimeGPT. Identifies outliers, level shifts, and trend breaks without model training. Use when identifying anomalies, outliers, or unusual patterns in time series. Trigger with "detect anomalies", "find outliers", "anomaly detection". |
| allowed-tools | Read,Write,Bash(python:*),Glob,Grep |
| version | 1.0.0 |
| author | Jeremy Longshore <jeremy@intentsolutions.io> |
| license | MIT |
Automatically detect and flag anomalies in time series data using TimeGPT.
This skill leverages TimeGPT's anomaly detection capabilities to identify outliers, level shifts, and trend breaks in time series data without requiring model training. It accepts CSV data, runs anomaly detection via the Nixtla API, and produces a detailed report with visualizations.
Required:
nixtla, pandas, matplotlib packagesEnvironment Variables:
NIXTLA_TIMEGPT_API_KEY: Your TimeGPT API keyInstallation:
pip install nixtla pandas matplotlib
Ensure your CSV file has the required Nixtla schema columns:
| Column | Type | Description |
|---|---|---|
unique_id | string | Series identifier |
ds | datetime | Timestamp |
y | numeric | Value to analyze |
export NIXTLA_TIMEGPT_API_KEY=your_api_key_here
Execute the detection script:
python {baseDir}/scripts/detect_anomalies.py --input your_data.csv
Available options:
--input, -i: Input CSV file (required)--output-csv, -o: Anomaly output CSV (default: anomalies.csv)--output-plot, -p: Visualization plot (default: anomalies_plot.png)--output-summary, -s: Summary text file (default: anomaly_summary.txt)The script generates three output files:
Error: NIXTLA_TIMEGPT_API_KEY not set
Solution: Run export NIXTLA_TIMEGPT_API_KEY=your_key
Error: CSV file missing required columns
Solution: Ensure CSV has unique_id, ds, and y columns
Error: No anomalies detected
Solution: This is valid output - data may have no anomalies
Error: Connection error to TimeGPT API
Solution: Check network connection and API key validity
Input (traffic.csv):
unique_id,ds,y
website_1,2024-01-01,1000
website_1,2024-01-02,1050
website_1,2024-01-03,300
website_1,2024-01-04,980
Command:
python {baseDir}/scripts/detect_anomalies.py --input traffic.csv
Output (anomalies.csv):
unique_id,ds,y,anomaly_type
website_1,2024-01-03,300,outlier
Input (sales.csv):
unique_id,ds,y
store_1,2023-12-28,50
store_1,2023-12-29,55
store_1,2023-12-30,60
store_1,2023-12-31,150
store_1,2024-01-01,145
Command:
python {baseDir}/scripts/detect_anomalies.py -i sales.csv -o sales_anomalies.csv
Output: Detects trend break at 2023-12-31
{baseDir}/scripts/detect_anomalies.py