mizcausevic-dev/fraud-click-filter

Purpose and principal workflow

Fraud Click Filter is a TypeScript demo control plane that presents sample traffic decisions, rules, threat signals, and spend-at-risk summaries to help growth and RevOps users reason about low-trust clicks. A browser user initiates the main workflow with an HTTP request: the Express app serves a rendered page or JSON response derived from in-memory sample data. The sampled implementation models decisions; it does not ingest or filter live ad traffic.

Components

HTTP entry and presentation

• Browser user — Path: null (external actor). Requests the documented pages and API routes to view traffic-integrity information.

• Express route application — src/app.ts. Defines HTML and JSON GET routes, calls renderers or service functions, and listens on loopback when run directly.

• Traffic dashboard renderer — src/services/render.ts. Builds the overview and grouped filter-lane, threat-signal, verification, and docs pages; its shown renderers obtain figures and records from the traffic service.

Domain data and projections

• Traffic sample dataset — src/data/sampleTraffic.ts. Defines event, rule, and threat-signal types and supplies the static records used by the demo.

• Traffic projections and payload — src/services/trafficFilterService.ts. Computes dashboard counts and spend at risk from events, exposes events/rules/signals/verification content, and assembles the combined sample payload.

Static output

• Prerendered site and API exporter — scripts/prerender.ts. Calls the same page renderers and service projections as the app, then writes HTML pages and JSON files under site/.

Relationships

• Browser user -> Express route application: sends HTTP GET requests to the documented routes.

• Express route application -> Traffic dashboard renderer: calls a page renderer for HTML routes.

• Traffic dashboard renderer -> Traffic projections and payload: reads summary, rule, signal, event, or verification projections to populate pages.

• Express route application -> Traffic projections and payload: calls the corresponding projection for JSON API routes.

• Traffic projections and payload -> Traffic sample dataset: reads the static events, rules, and threat signals; computes summary counts and spend from the events.

• Prerendered site and API exporter -> Traffic dashboard renderer: calls page renderers to produce static HTML.

• Prerendered site and API exporter -> Traffic projections and payload: calls projection functions to produce static JSON.

• Prerendered site and API exporter -> Local filesystem (external system, Path: null): writes generated HTML and JSON beneath site/.

Coverage limits

The source sample establishes an in-memory demo and its served and prerendered outputs, not a production traffic-ingestion or decision-enforcement pipeline. The README documents routes and product intent, but sampled code does not establish live integrations, persistent state, authentication, or downstream analytics/ad-platform connections. The renderer excerpt is partial, so details of unsampled page implementations are not inferred.