verdict-cli-0.1.6 is not a library.

verdict-cli — Rust-Based Data Validation for Pipelines and CI

Validate CSV files against a schema from the command line. A single static binary — no Python, no pip, no dependency conflicts. Installing pandas and pandera in CI typically takes 40–80 seconds; verdict-cli downloads in 1–3 seconds and runs immediately.

Built on verdict-core — a Rust validation engine with zero I/O overhead.

Installation

cargo install verdict-cli

Or download a pre-built binary for Linux, macOS, or Windows from the releases page.

Usage

verdict-cli data.csv schema.json
verdict-cli data.csv schema.yaml
verdict-cli data.csv schema.json --format text
verdict-cli data.csv schema.json --max-failed-samples 10

Schema format is detected from the file extension: .yaml / .yml → YAML, anything else → JSON.

Exit code 0 — all rules pass. Exit code 1 — at least one rule fails.

CI/CD integration

# GitHub Actions — no Rust or Python required
- uses: kkruglik/verdict@main
  with:
    csv: data/output.csv
    schema: data/schema.yaml

# pre-commit / shell script
verdict-cli data.csv schema.json || exit 1

Flags

Flag	Default	Description
`--format`	`json`	Output format: `json` or `text`
`--max-failed-samples`	`100`	Max failed row samples per rule in the report

Schema format

JSON:

{
  "columns": [
    { "name": "user_id", "dtype": "int", "constraints": [
      { "constraint": "not_null", "value": true },
      { "constraint": "unique",   "value": true }
    ]},
    { "name": "score", "dtype": "float", "constraints": [
      { "constraint": "between", "value": [0, 100] }
    ]},
    { "name": "country", "dtype": "str", "constraints": [
      { "constraint": "is_in", "value": ["US", "UK", "DE", "FR", "JP"] }
    ]},
    { "name": "created_date", "dtype": "date", "format": "%Y-%m-%d", "constraints": [
      { "constraint": "after", "value": "2020-01-01" }
    ]},
    { "name": "created_at", "dtype": "datetime", "format": "%Y-%m-%dT%H:%M:%S" }
  ]
}

YAML:

columns:
  - name: user_id
    dtype: int
    constraints:
      - constraint: not_null
        value: true
      - constraint: unique
        value: true
  - name: score
    dtype: float
    constraints:
      - constraint: between
        value: [0, 100]
  - name: country
    dtype: str
    constraints:
      - constraint: is_in
        value: ["US", "UK", "DE", "FR", "JP"]
  - name: created_date
    dtype: date
    format: "%Y-%m-%d"
    constraints:
      - constraint: after
        value: "2020-01-01"
  - name: created_at
    dtype: datetime
    format: "%Y-%m-%dT%H:%M:%S"

Supported dtypes: int, float, str, bool, date, datetime.

Date and DateTime columns accept an optional format string. If omitted, standard formats are tried (%Y-%m-%d for dates, %Y-%m-%dT%H:%M:%S and %Y-%m-%d %H:%M:%S for datetimes).

Columns without constraints are still required in the schema — they are loaded and type-checked but not validated against any rules.

Supported constraints

Constraint	Dtypes	Example value
`not_null`	all	`true`
`unique`	all	`true`
`gt` / `ge` / `lt` / `le`	int, float	`18`
`equal`	int, float, str	`"active"`
`between`	int, float	`[0, 100]`
`is_in`	int, float, str	`["US", "UK", "DE"]`
`matches_regex`	str	`"^[A-Z]{2}$"`
`contains`	str	`"@"`
`starts_with` / `ends_with`	str	`"prod_"`
`length_between`	str	`[2, 50]`
`after`	date, datetime	`"2020-01-01"`
`before`	date, datetime	`"2025-01-01"`
`between_dates`	date, datetime	`["2020-01-01", "2025-01-01"]`

Output

JSON (default):

{
  "passed": false,
  "total_rules": 3,
  "passed_count": 2,
  "failed_count": 1,
  "results": [
    {
      "column": "age",
      "constraint": "gt(18)",
      "passed": false,
      "failed_count": 2,
      "failed_values": [[3, "15"], [7, "12"]]
    }
  ]
}

Text (--format text):

Validation Report: FAILED (2/3 rules passed)
  FAIL: column 'age' — gt(18) — 2 values failed
    row 3: 15
    row 7: 12

License

MIT

verdict-cli 0.1.6