@buzzr/dfs-cli

Grade a DFS pick'em entry from two JSON files — no code required. This is a command-line wrapper around @buzzr/dfs-engine, with bounded inputs, effective-dated compatibility tables, policy metadata, and a full audit trail.

The operator-named built-ins are independent estimates: PrizePicks is experimental and partially verified; Underdog is experimental and unverified. Displayed entry terms are authoritative. Treat output as an audit aid, inspect policyStatus, policyVerification, payoutTable, confidence, sourceRefs, and explanationCodes, and confirm DNP, reboot, tie, rescue, void, and correction rulings with the operator.

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 |

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 | | Replay versioned engine regression cases | @buzzr/dfs-engine-test-vectors |

Links

• All-package API index
• Generated root-export reference
• Monorepo & full package family
• Issues

Compatibility and support

Node.js >= 22 is supported. Use the dfs-grade executable for the command line. Import the supported API from @buzzr/dfs-cli. Deep src/* and dist/* imports are unsupported.

Report vulnerabilities privately through SECURITY.md. The versioning and support policy defines the supported runtime and SemVer contract.

License

MIT