@buzzr/dfs-cli

Command-line wrapper around @buzzr/dfs-engine — grade a DFS entry from JSON fixtures without writing code.

Source

npm

Version: 5.0.0

Version published: 2 months ago

Weekly downloads: 3

Maintainers: 2

Weekly downloads

Created: 3 months ago

Source

@buzzr/dfs-cli

Grade a DFS pick'em entry from two JSON files — no code required. A command-line wrapper around @buzzr/dfs-engine, the settlement engine that grades PrizePicks/Underdog-style entries with book-accurate payout math and a full audit trail.

30-second quick start

npm install -g @buzzr/dfs-cli

Create entry.json (the slip you want to grade):

{
  "entryId": "entry-001",
  "bookId": "prizepicks",
  "playTypeId": "power",
  "stake": 10,
  "displayedMultiplier": 3,
  "legs": [
    {
      "legId": "leg-1",
      "playerName": "Jayson Tatum",
      "playerId": "athlete-a",
      "league": "NBA",
      "propType": "Points",
      "line": 26.5,
      "direction": "over",
      "actual": null,
      "status": "pending",
      "gameDate": "2026-05-07T00:00:00.000Z"
    }
  ]
}

And gamelogs.json (what the players actually did, keyed by legId):

{
  "leg-1": [
    {
      "date": "2026-05-07T00:00:00.000Z",
      "minutes": "38:00",
      "points": "31",
      "rebounds": "8",
      "assists": "5",
      "steals": "1",
      "blocks": "0",
      "turnovers": "2",
      "threeP": "4"
    }
  ]
}

Then grade it:

dfs-grade entry.json --gamelogs gamelogs.json

The full DfsSettlementResult prints to stdout as JSON — status, payout split, per-leg decisions with provider provenance, a validation report, and an audit trail:

{
  "entryId": "entry-001",
  "bookId": "prizepicks",
  "status": "won",
  "effectiveMultiplier": 3,
  "payout": { "total": 30, "withdrawable": 30, "bonus": 0 },
  "legs": [{ "legId": "leg-1", "status": "won", "actual": 31, "...": "..." }],
  "auditTrail": [{ "code": "settlement.won", "...": "..." }]
}

Exit code is 0 on success and 1 on any error (bad JSON, invalid entry, unknown book policy), so it drops straight into shell scripts and CI pipelines.

Programmatic API

The same grading path is exported as plain async functions:

import { runGrade, runGradeFromFiles } from '@buzzr/dfs-cli';

// From in-memory objects
const result = await runGrade({
  entry, // DfsEntryInput
  gameLogsByLegId: { 'leg-1': [gameLogRow] },
});

// Or straight from files
const fromFiles = await runGradeFromFiles({
  entryPath: './entry.json',
  gameLogsPath: './gamelogs.json',
});

Export	Purpose
`dfs-grade` (bin)	Grade an entry JSON against a gamelogs JSON, print settlement
`runGrade()`	Grade an in-memory `DfsEntryInput` against a `legId → gamelog[]` map
`runGradeFromFiles()`	Same, reading both inputs from file paths
`createDfsEngine` et al	Re-exported engine primitives for convenience

When to use this vs siblings

You want to…	Reach for
Grade entries from the shell, CI, or a cron job	this package
Grade entries inside a TypeScript/JavaScript app	`@buzzr/dfs-engine`
Render results in a UI	`@buzzr/dfs-react`
Build fixtures for your own tests	`@buzzr/dfs-testkit`
Verify your integration grades identically to Buzzr	`@buzzr/dfs-engine-test-vectors`