@platform/internal-logic

Mathematical and statistical calculation engine for ranked ballot voting systems. Implements 3-2-1 weighted Borda scoring, paired covariance and variance estimators, Z-score hypothesis testing for rank separation, Empirical Bayesian shrinkage for exposure regularization, and multi-factor rank skew anomaly detection.

---

Installation

The package is installed straight from this repository, pinned to a release tag:

npm install git+https://github.com/SeraphInteractive/Platform-Internal-Logic.git#v1.0.0

which records in package.json:

{
  "dependencies": {
    "@platform/internal-logic": "git+https://github.com/SeraphInteractive/Platform-Internal-Logic.git#v1.0.0"
  }
}

npm clones the tag, runs the prepare script (tsc) and installs only dist/, so consumers need git on the PATH but no build step of their own. The lockfile pins the exact commit.

Releasing

1. Bump version in package.json and merge to main (CI runs typecheck, tests and a trial install of the package from git).
2. Tag and push: git tag v1.1.0 && git push origin v1.1.0.
3. In consumers, bump the #v… ref and run npm install.

---

API Reference & Usage

import {
  validate_ballot,
  aggregate_scores,
  evaluate_rank_separation,
  calculate_bayesian_shrinkage,
  analyze_raid_risk,
  type Ballot,
} from '@platform/internal-logic';

// 1. Ballot validation (3-2-1 allocation, uniqueness, completeness)
const ballot: Ballot = {
  voterId: 'user_123',
  rank1: 'entry_a', // 3 points
  rank2: 'entry_b', // 2 points
  rank3: 'entry_c', // 1 point
};
const validation = validate_ballot(ballot);

// 2. Score aggregation & 6N conservation verification
const { scores, isConserved, leaderboard } = aggregate_scores(
  ['entry_a', 'entry_b', 'entry_c'],
  [ballot]
);

// 3. Paired Z-score rank separation test
const separation = evaluate_rank_separation(
  scores.get('entry_a')!,
  scores.get('entry_b')!,
  [ballot]
);

// 4. Empirical Bayesian shrinkage
const shrunkScore = calculate_bayesian_shrinkage(
  scores.get('entry_a')!,
  20, // Total entries in universe
  100 // Total ballots cast
);

// 5. Rank skew anomaly analysis
const telemetry = analyze_raid_risk(scores.get('entry_a')!);

---

Testing

Execute the test suite:

npm test