DataCheck

graph LR
    D["Data Files<br/>JSON / JSONL / CSV"] --> S["Schema<br/>(Inferred or Defined)"]
    S --> R["Rule Engine<br/>9 Rules + YAML Custom"]
    R --> A["Anomaly Detector<br/>IQR / Z-score"]
    A --> L["LLM Evaluator<br/>(Optional)"]
    L --> Rep["Quality Report<br/>MD / JSON / HTML"]
    Rep --> Fix["Auto Fix<br/>Dedup · PII · Trim"]
    Fix --> Diff["Report Diff<br/>Before vs After"]

    style R fill:#0969da,color:#fff,stroke:#0969da
    style A fill:#8b5cf6,color:#fff,stroke:#8b5cf6
    style Rep fill:#2da44e,color:#fff,stroke:#2da44e
    style Fix fill:#e5534b,color:#fff,stroke:#e5534b
    style D fill:#1a1a2e,color:#e0e0e0,stroke:#444
    style S fill:#1a1a2e,color:#e0e0e0,stroke:#444
    style L fill:#1a1a2e,color:#e0e0e0,stroke:#444
    style Diff fill:#1a1a2e,color:#e0e0e0,stroke:#444

# rules.yaml
rules:
  - field: instruction
    check: min_length
    value: 10
    severity: error

  - field: response
    check: max_length
    value: 10000
    severity: warning

  - field: category
    check: enum
    values: ["qa", "chat", "code", "math"]
    severity: error

knowlyr-datacheck check data.json --rules-file rules.yaml

knowlyr-datacheck check data.json    # 自动包含异常检测

# 1. 初次质检
knowlyr-datacheck check data.jsonl -o report_v1.json -f json

# 2. 自动修复（去重 + 去空白 + PII 脱敏）
knowlyr-datacheck fix data.jsonl -o fixed.jsonl --strip-pii

# 3. 再次质检
knowlyr-datacheck check fixed.jsonl -o report_v2.json -f json

# 4. 对比改进
knowlyr-datacheck diff report_v1.json report_v2.json

knowlyr-datacheck watch ./data/ --debounce 3 --ruleset sft

knowlyr-datacheck check data.json --ruleset llm
knowlyr-datacheck check data.json --ruleset llm --llm-provider openai

# Schema 推断
knowlyr-datacheck infer data.jsonl -o schema.json

# 批量检查（递归扫描所有数据文件）
knowlyr-datacheck check ./data/ --pattern "*.jsonl" -o report.html -f html

# 采样检查（大数据集）
knowlyr-datacheck check data.jsonl --sample 1000

pip install knowlyr-datacheck

pip install knowlyr-datacheck[stats]    # 统计分析 (numpy, scipy)
pip install knowlyr-datacheck[mcp]      # MCP 服务器
pip install knowlyr-datacheck[llm]      # LLM 智能检查
pip install knowlyr-datacheck[yaml]     # YAML 规则配置
pip install knowlyr-datacheck[watch]    # Watch 模式
pip install knowlyr-datacheck[all]      # 全部功能

# 基础检查（支持 JSON / JSONL / CSV）
knowlyr-datacheck check data.json

# 指定 Schema + 输出报告
knowlyr-datacheck check data.json -s schema.json -o report.md

# HTML 报告
knowlyr-datacheck check data.json -o report.html -f html

# CI 集成：设定通过率阈值
knowlyr-datacheck check data.json --threshold 0.9 --strict

# 数据修复
knowlyr-datacheck fix data.jsonl -o fixed.jsonl --strip-pii

from datacheck import DataChecker, QualityReport

checker = DataChecker()
result = checker.check_file("data.json", schema_path="schema.json")

report = QualityReport(result)
report.print_summary()
report.save("./report.md")

# 验证 DataRecipe 分析结果中的合成数据
knowlyr-datacheck validate ./analysis_output/my_dataset/
knowlyr-datacheck validate ./analysis_output/my_dataset/ -d custom_data.json

from datacheck.anomaly import detect_anomalies

anomalies = detect_anomalies(samples)
for field, info in anomalies.items():
    print(f"{field}: {info['outlier_count']} outliers, range [{info['bounds']['lower']}, {info['bounds']['upper']}]")

{
  "mcpServers": {
    "knowlyr-datacheck": {
      "command": "uv",
      "args": ["--directory", "/path/to/data-check", "run", "python", "-m", "datacheck.mcp_server"]
    }
  }
}

- uses: actions/setup-python@v5
  with:
    python-version: '3.12'
- run: pip install knowlyr-datacheck
- run: knowlyr-datacheck check data.json --threshold 0.9 --strict

graph LR
    Radar["Radar<br/>Discovery"] --> Recipe["Recipe<br/>Analysis"]
    Recipe --> Synth["Synth<br/>Generation"]
    Recipe --> Label["Label<br/>Annotation"]
    Synth --> Check["Check<br/>Quality"]
    Label --> Check
    Check --> Audit["Audit<br/>Model Audit"]
    Crew["Crew<br/>Deliberation Engine"]
    Agent["Agent<br/>RL Framework"]
    ID["ID<br/>Identity Runtime"]
    Crew -.->|能力定义| ID
    ID -.->|身份 + 记忆| Crew
    Crew -.->|轨迹 + 奖励| Agent
    Agent -.->|优化策略| Crew

    style Check fill:#0969da,color:#fff,stroke:#0969da
    style Crew fill:#2da44e,color:#fff,stroke:#2da44e
    style Agent fill:#8b5cf6,color:#fff,stroke:#8b5cf6
    style ID fill:#e5534b,color:#fff,stroke:#e5534b
    style Radar fill:#1a1a2e,color:#e0e0e0,stroke:#444
    style Recipe fill:#1a1a2e,color:#e0e0e0,stroke:#444
    style Synth fill:#1a1a2e,color:#e0e0e0,stroke:#444
    style Label fill:#1a1a2e,color:#e0e0e0,stroke:#444
    style Audit fill:#1a1a2e,color:#e0e0e0,stroke:#444

git clone https://github.com/liuxiaotong/data-check.git
cd data-check
pip install -e ".[all,dev]"
pytest