Candidate Fraud Risk Assessment

A standalone, single-file HTML tool for assessing candidate fraud risk — designed to detect patterns consistent with DPRK IT worker fraud and similar hiring threats.

---

---

Usage

Open Candidate_Risk_Dashboard.html in any browser. No server, no install, no build step required.

Modes

• Example mode — displays a pre-filled CRITICAL-risk case based on a confirmed DPRK IT worker fraud scenario. Inputs are read-only.
• Assess mode — blank scorecard with editable candidate name/role fields, a Reset button, and a Copy Review Link button that encodes your full assessment as a URL-safe base64 string for sharing.
• Panel View — aggregate multiple reviewers' assessments into a composite score. Paste review links from each interviewer to combine them. Shows per-reviewer totals, per-category averages, and flags categories with MODERATE SPREAD or HIGH DISAGREEMENT across reviewers.

Switching between Example and Assess modes preserves each scorecard's state independently.

Panel Review Workflow

1. Each interviewer completes their own assessment in Assess mode and clicks Copy Review Link.
2. A coordinator opens Panel View and pastes each reviewer's link.
3. The panel URL auto-updates in the browser address bar — bookmark or share it to preserve the full panel state.
4. Categories where reviewers disagree are flagged automatically. Categories marked "Not Assessed" by a reviewer are excluded from that reviewer's averages.

Scoring

Sixteen categories span four hiring stages. Each category includes a Not Assessed / N/A option.

Stage	Categories	Max Points
Prescreen	Digital Footprint, Resume Authenticity, Phone / Contact Integrity, Geographic Consistency, Timezone Behavior Patterns, Account / Identity Duplication, Third-Party Handler	82
Interview	Interview Signals, Technical Screen Signals, Reference Integrity	41
Offer	Compensation Behavior, Employer Verification	23
Background Check	Identity Documentation, Payment / Financial Flags, Hardware / Onboarding, Identity Continuity	50

Risk tiers (0–196 total points):

Score	Level	Action
0–30	LOW	Proceed normally. Standard monitoring.
31–70	MODERATE	Proceed with caution. Additional OSINT verification. Flag for CISO awareness.
71–120	HIGH	Do NOT extend offer without CISO approval. Full OSINT + live ID verification required.
121+	CRITICAL	REJECT. CISO + Legal sign-off required. Preserve evidence. Initiate reporting.

Development

npm install            # install dev dependencies (first time only)
pre-commit install     # activate pre-commit hook (first time only)

npm run lint           # run all linters (HTML + JS + Actions)
npm run lint:html      # HTMLHint only
npm run lint:js        # ESLint only (lints inline <script> via eslint-plugin-html)
npm run lint:actions   # actionlint only (GitHub Actions workflows)

The pre-commit hook runs all three linters on every commit (actionlint for .yml files, HTMLHint + ESLint for .html files).

Security

• No data is sent anywhere — entirely client-side, no network calls from application code.
• CSP restricts scripts to cdnjs.cloudflare.com (React 18 with SRI hashes) and inline scripts only.
• URL sharing uses btoa/atob builtins — no new CDN dependencies, no CSP changes required.
• Cache-control headers, no-referrer, noindex, and Permissions-Policy are intentional — do not remove them.

License

MIT License — see LICENSE. Copyright (c) 2026 Colorado = Security.