datadrivenconstruction

Data Anomaly Detector

資料分析v2.1.0
作者:datadrivenconstruction更新於 2026-05-11共 2 個版本 · 最新版本釋出於 2026-02-15data-anomaly-detector@clawhub_datadrivenconstruction/data-anomaly-detector
資料洞察指標監控#Schedule

據 AlphaMoat 統計,截至 2026-10,Data Anomaly Detector 累計下載 6,746 次、獲 2 個星標,按累計下載量在本次參與排名的 10,914 個 AI 技能中排名第 4,220 位,在「資料分析」分類(915 個)中排名第 397 位。

列表預設按累計下載量排序,也可按星標排序;詳情頁排名按累計下載量計算。

累計下載
6746
星標
2
全站下載量排名
#4,220
全站前 39% · 「資料分析」前 44%
平台累計下載星標評論
ClawHub ↗51781-
SkillHub ↗67462—

各平台的計數口徑不同,統計時間不一致。

這個技能是做什麼的

主文涉及成本、進度、生產率和資料質量檢查,包含統計及機器學習方法。專案資料可能具有商業敏感性,異常只是調查線索,不證明欺詐或因果;核實資料來源、單位和現場背景。

資料釋出於 2026-10-27

可以用來做什麼

  • 篩查成本或進度資料的異常項
  • 檢查重複記錄和規則偏差
  • 比較生產率資料中的趨勢變化

什麼時候用

  • 需要對提供的建築專案資料做異常篩查時
  • 核實資料口徑和現場背景後再將異常作為調查結論時

示例提示詞

以下是根據技能文件生成的示例,不代表實際執行結果。

檢查這些專案資料中的異常候選,說明篩查依據和可能誤報,不要將其直接歸因於欺詐。

月度資料趨勢

累計下載6746
3131507370142026-03 · 累計下載 33992026-05 · 累計下載 44822026-06 · 累計下載 45962026-07 · 累計下載 48242026-08 · 累計下載 49492026-09 · 累計下載 53002026-10 · 累計下載 674626-0326-0926-10
星標2
1222026-03 · 星標 12026-05 · 星標 22026-06 · 星標 22026-07 · 星標 22026-08 · 星標 22026-09 · 星標 22026-10 · 星標 226-0326-0926-10

2026-09 起統計方法調整:同名技能按作者分別統計,並覆蓋全部技能。圖中虛線為調整前的月份,與之後的資料不直接比較,也不計算跨越調整的增長。

月份累計下載環比星標
2026-106746—2
2026-095300—2
2026-084949+2.6%2
2026-074824+5.0%2
2026-064596+2.5%2
2026-054482—2
2026-033399—1

SKILL.md 原文

以下是作者釋出的原始檔案,按純文字顯示,本站未做修改或稽核;其中的連結不可點選,命令請核實後再執行。抓取於 2026-10-23。

19 KB · ClawHub v2.1.0
---
name: "data-anomaly-detector"
description: "Detect anomalies and outliers in construction data: unusual costs, schedule variances, productivity spikes. Statistical and ML-based detection methods."
homepage: "https://datadrivenconstruction.io"
metadata: {"openclaw": {"emoji": "✔️", "os": ["darwin", "linux", "win32"], "homepage": "https://datadrivenconstruction.io", "requires": {"bins": ["python3"]}}}
---
# Data Anomaly Detector for Construction

## Overview

展開全部(共 514 行)
Detect unusual patterns, outliers, and anomalies in construction data. Identify cost overruns, schedule delays, productivity issues, and data quality problems before they impact projects.

## Business Case

Construction data often contains anomalies that indicate:
- Cost estimate errors or fraud
- Schedule logic issues
- Productivity problems
- Data entry mistakes
- Equipment or material issues

Early detection prevents costly corrections and project delays.

## Technical Implementation

```python
from dataclasses import dataclass, field
from typing import List, Dict, Any, Optional, Tuple
from enum import Enum
import pandas as pd
import numpy as np
from datetime import datetime
from scipy import stats

class AnomalyType(Enum):
    OUTLIER = "outlier"
    PATTERN_BREAK = "pattern_break"
    MISSING_SEQUENCE = "missing_sequence"
    DUPLICATE = "duplicate"
    IMPOSSIBLE_VALUE = "impossible_value"
    TREND_DEVIATION = "trend_deviation"

class AnomalySeverity(Enum):
    CRITICAL = "critical"
    HIGH = "high"
    MEDIUM = "medium"
    LOW = "low"

@dataclass
class Anomaly:
    id: str
    anomaly_type: AnomalyType
    severity: AnomalySeverity
    field: str
    value: Any
    expected_range: Optional[Tuple[float, float]] = None
    description: str = ""
    row_index: Optional[int] = None
    detection_method: str = ""
    confidence: float = 0.0
    suggested_action: str = ""

@dataclass
class AnomalyReport:
    source: str
    detected_at: datetime
    total_records: int
    anomalies: List[Anomaly]
    summary: Dict[str, int]

class ConstructionAnomalyDetector:
    """Detect anomalies in construction data."""

    # Construction-specific thresholds
    COST_THRESHOLDS = {
        'concrete_per_cy': (200, 800),
        'steel_per_ton': (1500, 4000),
        'labor_per_hour': (25, 150),
        'overhead_percentage': (5, 25),
        'contingency_percentage': (3, 20),
    }

    SCHEDULE_THRESHOLDS = {
        'max_activity_duration': 365,  # days
        'max_lag': 30,  # days
        'min_productivity': 0.1,
        'max_productivity': 10.0,
    }

    def __init__(self):
        self.anomalies: List[Anomaly] = []
        self.detection_history: List[AnomalyReport] = []

    def detect_cost_anomalies(self, df: pd.DataFrame, cost_column: str,
                              group_by: str = None) -> List[Anomaly]:
        """Detect anomalies in cost data."""
        anomalies = []

        # Statistical outlier detection (IQR method)
        Q1 = df[cost_column].quantile(0.25)
        Q3 = df[cost_column].quantile(0.75)
        IQR = Q3 - Q1
        lower_bound = Q1 - 1.5 * IQR
        upper_bound = Q3 + 1.5 * IQR

        outliers = df[(df[cost_column] < lower_bound) | (df[cost_column] > upper_bound)]

        for idx, row in outliers.iterrows():
            value = row[cost_column]
            severity = AnomalySeverity.HIGH if abs(value - df[cost_column].median()) > 3 * IQR else AnomalySeverity.MEDIUM

            anomalies.append(Anomaly(
                id=f"COST-{idx}",
                anomaly_type=AnomalyType.OUTLIER,
                severity=severity,
                field=cost_column,
                value=value,
                expected_range=(lower_bound, upper_bound),
                description=f"Cost value {value:,.2f} outside expected range",
                row_index=idx,
                detection_method="IQR",
                confidence=0.95,
                suggested_action="Review cost estimate for errors"
            ))

        # Negative cost check
        negatives = df[df[cost_column] < 0]
        for idx, row in negatives.iterrows():
            anomalies.append(Anomaly(
                id=f"COST-NEG-{idx}",
                anomaly_type=AnomalyType.IMPOSSIBLE_VALUE,
                severity=AnomalySeverity.CRITICAL,
                field=cost_column,
                value=row[cost_column],
                expected_range=(0, None),
                description="Negative cost value detected",
                row_index=idx,
                detection_method="Business Rule",
                confidence=1.0,
                suggested_action="Correct data entry error or investigate credit"
            ))

        # Group-based anomalies (if grouped)
        if group_by and group_by in df.columns:
            group_stats = df.groupby(group_by)[cost_column].agg(['mean', 'std'])

            for group_name, stats in group_stats.iterrows():
                group_data = df[df[group_by] == group_name]
                z_scores = np.abs((group_data[cost_column] - stats['mean']) / stats['std'])

                for idx, z in z_scores.items():
                    if z > 3:
                        anomalies.append(Anomaly(
                            id=f"COST-GROUP-{idx}",
                            anomaly_type=AnomalyType.OUTLIER,
                            severity=AnomalySeverity.MEDIUM,
                            field=cost_column,
                            value=df.loc[idx, cost_column],
                            description=f"Unusual cost for group {group_name} (z-score: {z:.2f})",
                            row_index=idx,
                            detection_method="Z-Score by Group",
                            confidence=min(z / 5, 1.0)
                        ))

        return anomalies

    def detect_schedule_anomalies(self, df: pd.DataFrame) -> List[Anomaly]:
        """Detect anomalies in schedule data."""
        anomalies = []

        # Check for required columns
        required = ['start_date', 'end_date']
        if not all(col in df.columns for col in required):
            return anomalies

        # Convert dates
        df['start_date'] = pd.to_datetime(df['start_date'])
        df['end_date'] = pd.to_datetime(df['end_date'])

        # Calculate duration
        df['duration'] = (df['end_date'] - df['start_date']).dt.days

        # Negative duration (end before start)
        negative_duration = df[df['duration'] < 0]
        for idx, row in negative_duration.iterrows():
            anomalies.append(Anomaly(
                id=f"SCHED-NEG-{idx}",
                anomaly_type=AnomalyType.IMPOSSIBLE_VALUE,
                severity=AnomalySeverity.CRITICAL,
                field="duration",
                value=row['duration'],
                description="End date before start date",
                row_index=idx,
                detection_method="Business Rule",
                confidence=1.0,
                suggested_action="Correct dates"
            ))

        # Extremely long durations
        long_tasks = df[df['duration'] > self.SCHEDULE_THRESHOLDS['max_activity_duration']]
        for idx, row in long_tasks.iterrows():
            anomalies.append(Anomaly(
                id=f"SCHED-LONG-{idx}",
                anomaly_type=AnomalyType.OUTLIER,
                severity=AnomalySeverity.MEDIUM,
                field="duration",
                value=row['duration'],
                expected_range=(0, self.SCHEDULE_THRESHOLDS['max_activity_duration']),
                description=f"Task duration {row['duration']} days exceeds threshold",
                row_index=idx,
                detection_method="Threshold",
                confidence=0.9,
                suggested_action="Review if task should be broken down"
            ))

        # Zero duration non-milestones
        if 'is_milestone' in df.columns:
            zero_duration = df[(df['duration'] == 0) & (~df['is_milestone'])]
            for idx, row in zero_duration.iterrows():
                anomalies.append(Anomaly(
                    id=f"SCHED-ZERO-{idx}",
                    anomaly_type=AnomalyType.IMPOSSIBLE_VALUE,
                    severity=AnomalySeverity.HIGH,
                    field="duration",
                    value=0,
                    description="Zero duration task that is not a milestone",
                    row_index=idx,
                    detection_method="Business Rule",
                    confidence=1.0,
                    suggested_action="Add duration or mark as milestone"
                ))

        return anomalies

    def detect_productivity_anomalies(self, df: pd.DataFrame,
                                      quantity_col: str,
                                      hours_col: str) -> List[Anomaly]:
        """Detect productivity anomalies."""
        anomalies = []

        # Calculate productivity
        df['productivity'] = df[quantity_col] / df[hours_col].replace(0, np.nan)

        # Use Modified Z-Score (more robust for skewed data)
        median = df['productivity'].median()
        mad = np.abs(df['productivity'] - median).median()
        modified_z = 0.6745 * (df['productivity'] - median) / mad

        outliers = df[np.abs(modified_z) > 3.5]

        for idx, row in outliers.iterrows():
            prod = row['productivity']
            z = modified_z.loc[idx]

            severity = AnomalySeverity.HIGH if abs(z) > 5 else AnomalySeverity.MEDIUM
            direction = "high" if z > 0 else "low"

            anomalies.append(Anomaly(
                id=f"PROD-{idx}",
                anomaly_type=AnomalyType.OUTLIER,
                severity=severity,
                field="productivity",
                value=prod,
                description=f"Unusually {direction} productivity: {prod:.2f} units/hour",
                row_index=idx,
                detection_method="Modified Z-Score",
                confidence=min(abs(z) / 7, 1.0),
                suggested_action=f"Investigate {direction} productivity cause"
            ))

        return anomalies

    def detect_time_series_anomalies(self, df: pd.DataFrame,
                                      date_col: str,
                                      value_col: str,
                                      window: int = 7) -> List[Anomaly]:
        """Detect anomalies in time series data (e.g., daily costs, progress)."""
        anomalies = []

        df = df.sort_values(date_col).copy()
        df['rolling_mean'] = df[value_col].rolling(window=window, center=True).mean()
        df['rolling_std'] = df[value_col].rolling(window=window, center=True).std()

        # Points outside 2 standard deviations from rolling mean
        df['z_score'] = (df[value_col] - df['rolling_mean']) / df['rolling_std']

        outliers = df[np.abs(df['z_score']) > 2].dropna()

        for idx, row in outliers.iterrows():
            anomalies.append(Anomaly(
                id=f"TS-{idx}",
                anomaly_type=AnomalyType.TREND_DEVIATION,
                severity=AnomalySeverity.MEDIUM if abs(row['z_score']) < 3 else AnomalySeverity.HIGH,
                field=value_col,
                value=row[value_col],
                expected_range=(
                    row['rolling_mean'] - 2 * row['rolling_std'],
                    row['rolling_mean'] + 2 * row['rolling_std']
                ),
                description=f"Value deviates from {window}-day trend",
                row_index=idx,
                detection_method="Rolling Z-Score",
                confidence=min(abs(row['z_score']) / 4, 1.0)
            ))

        return anomalies

    def detect_duplicate_anomalies(self, df: pd.DataFrame,
                                   key_columns: List[str]) -> List[Anomaly]:
        """Detect duplicate records."""
        anomalies = []

        duplicates = df[df.duplicated(subset=key_columns, keep=False)]

        if len(duplicates) > 0:
            dup_groups = duplicates.groupby(key_columns).size()
            for keys, count in dup_groups.items():
                anomalies.append(Anomaly(
                    id=f"DUP-{hash(str(keys)) % 10000}",
                    anomaly_type=AnomalyType.DUPLICATE,
                    severity=AnomalySeverity.HIGH,
                    field=str(key_columns),
                    value=keys,
                    description=f"Found {count} duplicate records for {keys}",
                    detection_method="Exact Match",
                    confidence=1.0,
                    suggested_action="Review and remove duplicates"
                ))

        return anomalies

    def detect_sequence_gaps(self, df: pd.DataFrame, sequence_col: str) -> List[Anomaly]:
        """Detect gaps in sequential data (invoice numbers, PO numbers, etc.)."""
        anomalies = []

        # Extract numeric part if mixed format
        df['seq_num'] = pd.to_numeric(
            df[sequence_col].astype(str).str.extract(r'(\d+)')[0],
            errors='coerce'
        )

        sorted_seq = df['seq_num'].dropna().sort_values()
        expected = range(int(sorted_seq.min()), int(sorted_seq.max()) + 1)
        actual = set(sorted_seq.astype(int))
        missing = set(expected) - actual

        if missing:
            # Group consecutive missing numbers
            missing_ranges = []
            sorted_missing = sorted(missing)
            start = sorted_missing[0]
            end = start

            for num in sorted_missing[1:]:
                if num == end + 1:
                    end = num
                else:
                    missing_ranges.append((start, end))
                    start = num
                    end = num
            missing_ranges.append((start, end))

            for start, end in missing_ranges:
                range_str = str(start) if start == end else f"{start}-{end}"
                anomalies.append(Anomaly(
                    id=f"SEQ-{start}",
                    anomaly_type=AnomalyType.MISSING_SEQUENCE,
                    severity=AnomalySeverity.MEDIUM,
                    field=sequence_col,
                    value=range_str,
                    description=f"Missing sequence number(s): {range_str}",
                    detection_method="Sequence Analysis",
                    confidence=1.0,
                    suggested_action="Investigate missing numbers"
                ))

        return anomalies

    def run_full_detection(self, df: pd.DataFrame, config: Dict) -> AnomalyReport:
        """Run all applicable anomaly detection methods."""
        all_anomalies = []

        # Cost anomalies
        if 'cost_columns' in config:
            for col in config['cost_columns']:
                if col in df.columns:
                    all_anomalies.extend(
                        self.detect_cost_anomalies(df, col, config.get('group_by'))
                    )

        # Schedule anomalies
        if 'start_date' in df.columns and 'end_date' in df.columns:
            all_anomalies.extend(self.detect_schedule_anomalies(df))

        # Productivity
        if 'quantity_col' in config and 'hours_col' in config:
            all_anomalies.extend(
                self.detect_productivity_anomalies(
                    df, config['quantity_col'], config['hours_col']
                )
            )

        # Duplicates
        if 'key_columns' in config:
            all_anomalies.extend(
                self.detect_duplicate_anomalies(df, config['key_columns'])
            )

        # Sequence gaps
        if 'sequence_column' in config:
            all_anomalies.extend(
                self.detect_sequence_gaps(df, config['sequence_column'])
            )

        # Create summary
        summary = {}
        for a in all_anomalies:
            key = f"{a.anomaly_type.value}_{a.severity.value}"
            summary[key] = summary.get(key, 0) + 1

        report = AnomalyReport(
            source=config.get('source_name', 'Unknown'),
            detected_at=datetime.now(),
            total_records=len(df),
            anomalies=all_anomalies,
            summary=summary
        )

        self.detection_history.append(report)
        return report

    def generate_report(self, report: AnomalyReport) -> str:
        """Generate markdown anomaly report."""
        lines = [f"# Anomaly Detection Report", ""]
        lines.append(f"**Source:** {report.source}")
        lines.append(f"**Detected At:** {report.detected_at.strftime('%Y-%m-%d %H:%M')}")
        lines.append(f"**Total Records:** {report.total_records:,}")
        lines.append(f"**Anomalies Found:** {len(report.anomalies)}")
        lines.append("")

        # Summary by severity
        lines.append("## Summary by Severity")
        for severity in AnomalySeverity:
            count = sum(1 for a in report.anomalies if a.severity == severity)
            if count > 0:
                lines.append(f"- **{severity.value.upper()}:** {count}")
        lines.append("")

        # Critical anomalies first
        critical = [a for a in report.anomalies if a.severity == AnomalySeverity.CRITICAL]
        if critical:
            lines.append("## Critical Anomalies")
            for a in critical:
                lines.append(f"\n### {a.id}")
                lines.append(f"- **Type:** {a.anomaly_type.value}")
                lines.append(f"- **Field:** {a.field}")
                lines.append(f"- **Value:** {a.value}")
                lines.append(f"- **Description:** {a.description}")
                lines.append(f"- **Action:** {a.suggested_action}")

        # All anomalies table
        lines.append("\n## All Anomalies")
        lines.append("| ID | Type | Severity | Field | Description |")
        lines.append("|-----|------|----------|-------|-------------|")
        for a in report.anomalies[:50]:
            lines.append(f"| {a.id} | {a.anomaly_type.value} | {a.severity.value} | {a.field} | {a.description[:50]} |")

        if len(report.anomalies) > 50:
            lines.append(f"\n*... and {len(report.anomalies) - 50} more anomalies*")

        return "\n".join(lines)
```

## Quick Start

```python
import pandas as pd

# Load data
df = pd.read_excel("project_costs.xlsx")

# Initialize detector
detector = ConstructionAnomalyDetector()

# Run detection
config = {
    'source_name': 'Project Costs Q1 2026',
    'cost_columns': ['total_cost', 'labor_cost', 'material_cost'],
    'group_by': 'cost_code',
    'key_columns': ['project_id', 'cost_code', 'date'],
    'sequence_column': 'invoice_number'
}

report = detector.run_full_detection(df, config)

# Generate report
print(detector.generate_report(report))

# Get critical anomalies for immediate action
critical = [a for a in report.anomalies if a.severity == AnomalySeverity.CRITICAL]
print(f"\n{len(critical)} critical anomalies require immediate attention")
```

## Dependencies

```bash
pip install pandas numpy scipy
```

## Resources

- **Statistical Methods**: IQR, Z-Score, Modified Z-Score
- **Construction Benchmarks**: RSMeans, ENR indices

「資料分析」分類下的熱門技能

全能金融爬蟲資料分析

中國金融機構資料爬取與分析綜合 Skill v12.0.0。2334家/43大類機構名單、60+財經媒體源輿情、 A股報告/公告/研報、60+種查詢多源金融資料(自動降級+快取)、爬取排程與網頁操作增強、 即時資料中樞、24 技術指標引擎、多源證據融合、110 MCP 工具 + 10 Resource + 14 Prompt。 🆕 v12.0.0 四條線(資訊獲取 / 打包下載 / 資料分析 / 登入): 編排:gather_info() 一次跑完"路由→抓取→去佔位→質量分→能力邊界→回執→打包", 8 類意圖各帶問法、輸入輸出、典型耗時與 not_for(說清哪件事做不了)。 誠實:反爬佔位條目一律剔除並記賬,空結果報 E_SOURCE_EMPTY,不把假資料判成 B 級成功。 交付:包內自帶 manifest.json(逐檔案 sha256),verify 查得出改動/缺失/多塞, 批次下載單條失敗不中斷整批,失敗清單可直接只重下那幾條。 登入:Cookie 由使用者匯出或掃碼接管,檔案化 + 健康度四檔,對外輸出全脫敏(值永不回顯)。 何時觸發:爬取機構、輿情爬取、東方財富、公告下載、即時行情、技術指標、統一查詢、 交付包、批次重下、我已經登入了、登入態還有效嗎(完整觸發詞見下方 auto_trigger)。

↓ 77萬★ 288user_9d5a2a39
鋼聯AI - 智慧問數資料分析

大宗商品價格資料查詢技能。用於查詢大宗商品期現貨價格、宏觀經濟指標、產業鏈供需資料、進出口庫存資料、金融市場資料等。當用戶詢問:(1) 期貨/現貨價格查詢,如"螺紋鋼價格"、"銅期貨收盤價"; (2) 歷史行情或趨勢分析;(3) 區域或企業價格對比;(4) 宏觀經濟指標,如GDP、CPI;(5) 產業鏈上下游資料;(6) 多市場金融資料(期權、股票、債券、外匯);(7) 結算確認單生成時觸發此技能。

↓ 70.5萬★ 6u_9ed3401d
smart-charts資料分析

32種圖表,3 個主題,2步生成: 1.上傳資料 — 將 CSV / Excel / JSON 等格式的資料檔案拖入對話方塊; 2.檢視結果 — 互動式圖表(HTML),可儲存圖表為圖片.

↓ 60.5萬★ 135user_5b28ea14
AdMapix資料分析

AdMapix 的取數層:把廣告與應用相關的查詢轉成介面呼叫,只回原始 JSON。

↓ 57.5萬★ 304fly0pants
論文配圖一鍵生成資料分析

論文配圖一鍵生成:一條命令從 JSON/CSV/Excel 生成頂刊級學術圖表,科研繪圖 與期刊配圖全場景覆蓋。40 種圖表(森林圖/KM 生存曲線(自動風險表)/競爭風險累計發生率/ROC/校準 曲線/決策曲線 DCA/列線圖/RCS 非線性劑量反應/亞組分析森林圖/RMST 限制平均生存時間/Galbraith 徑向圖/火山圖/雨雲圖/山脊圖/啞鈴圖/游泳圖(逐例時長+里程碑)/ 腫瘤緩解瀑布圖/UpSet 交集/PRISMA 流程圖/漏斗/BA 一致性/PCA/熱圖/箱線/小提琴/ 組合圖/Cox 多因素迴歸森林圖等),9 套配色(含色盲安全 Okabe-Ito 與期刊色板),9 種期刊投稿預設, YAML 流水線一條命令出全套論文圖,一次渲染多格式匯出(TIFF/PNG/PDF), PDF 重疊門禁,600dpi。純本地執行,資料不出機,中文零配置。適用:寫論文/ 投稿要配圖、把資料變成可投稿圖表、批次重出全套圖。不適用:文字潤色/翻譯/ 改寫(請用寫作類工具);商業報表/儀表盤/海報裝飾;無資料支撐的純示意圖。

↓ 42.1萬★ 83user_481397b5
smart-charts資料分析

Intelligent chart generation and data analysis skill. Reads user-supplied data files (CSV/Excel/JSON), analyzes data characteristics with LLM assistance, auto-recommends and generates interactive ECharts visualizations. Use when the user asks to analyze data, generate charts, create visualizations, or work with tabular data files.

↓ 25.7萬★ 1user_814dbe54
股票/基金/期貨量化分析工具資料分析

全球智慧投研工具(stock-researcher)v13.0:多源資料 + 機率預測 + 可匯出報告。 核心能力: - 多源資料路由:按市場/類別選源,逐級降級;每個結果標註來源與時效等級;源不可用時明確標 unavailable - 全資產量化:股票/ETF/場外基金/指數/期貨(期限結構、基差、持倉、折溢價)多因子分析 - 機率預測:1d/5d/1M/3M/6M 經驗分佈 + 分位數 + 方向機率 + 區間,kNN 鄰域收縮融合全域性分佈; 樣本不足自動棄權,可按歷史命中樣本(同桶)做機率校準 - 證據鏈簡報:每條結論繫結來源、時間與資料質量,衝突訊號保留而不抹平 - 報告匯出:Word/PPT/PDF/Markdown/JSON/純文字,純標準庫渲染 + 可選第三方引擎 - MCP Server 24 工具,Claude Code 可直接呼叫;知識庫 12 介面卡雙向同步 三條鐵律: - 資料不裝:拿不到就報不可用,不拿 0 價、不拿空表冒充 - 預測不裝:機率來自歷史相似條件的經驗分佈,樣本不足就說不足,不硬給方向 - 評分≠盈利保證:任何結論都帶免責宣告

↓ 19.2萬★ 147user_9d5a2a39
Tavily Search資料分析

透過 Tavily 搜尋公開網頁,並回傳含連結與摘要的結構化結果。

↓ 16.5萬★ 286Xiaoran Li

datadrivenconstruction 的其他技能

資料月份:2026-10 · 下載/星標數聚合自公開技能註冊庫(ClawHub、SkillHub),按月更新。口徑說明見資料方法。