Internal Tool

Vigil

Always watching. Never blinking.

Explore Features

Screenshots

See Vigil in Action

Features

Know Before Your Users Do

Continuous endpoint monitoring with sub-second WebSocket alerts, latency tracking, incident management, and uptime reporting. When something breaks at 3 AM, Vigil is already on it.

Health Monitoring

Endpoints, Latency, and Uptime

01

Automated Endpoint Polling

Configurable check intervals per site. HTTP status validation, response time measurement, and content verification. APScheduler-based polling engine.

02

Latency Tracking

Millisecond-resolution response time history. Sparkline visualizations on dashboard cards. Detailed timeseries charts in site detail view.

03

Uptime Statistics

Rolling uptime percentages across 24-hour, 7-day, 30-day, and 90-day windows. Per-site and aggregate fleet uptime on the dashboard.

04

Health Status Classification

Three-state health model: healthy (green), degraded (yellow), and down (red). Degraded status triggered by slow response times before full failure.

Incident Management

Alerts, Incidents, and Resolution

05

Automatic Incident Creation

Incidents created automatically when health checks fail. Tracks start time, affected service, and current status. Auto-resolves when the service recovers.

06

Alarm Banner

Red banner across the top of the dashboard when any service is down. Shows count of affected sites. Mutable for acknowledged incidents.

07

WebSocket Live Updates

Real-time check results and incident broadcasts via WebSocket. Dashboard updates instantly without polling. No refresh needed.

08

Incident History

Full incident timeline per site with start, end, and duration. Filter by status (active, resolved). Incident table in the site detail modal.

Dashboard

Card View and List View

09

Card View

Grid of site cards showing name, status indicator, uptime percentage, latency sparkline, and last check time. Color-coded borders for health state.

10

List View

Tabular view with sortable columns for name, URL, status, uptime, latency, and last check. Toggle between card and list from the header.

11

Site Detail Modal

Click any site to open a detail panel with latency chart (Recharts), uptime stats across all time windows, incident table, sub-checks, and settings.

12

Status Summary

Aggregated counts at the top: total sites, healthy, degraded, and down. Fleet-wide uptime percentage for the last 24 hours.

Accessibility

Built for Everyone

WCAG 2.1 AA Compliance

4.5:1 contrast for body text, 3:1 for large text and UI components, in both light and dark themes.

Keyboard Navigation

Every interaction reachable via keyboard. Logical tab order, visible focus indicators, Escape-to-dismiss for modals.

Screen Reader Support

VoiceOver, NVDA, and JAWS tested. Semantic HTML, ARIA labels, live regions for dynamic updates.

Reduced Motion

Respects prefers-reduced-motion. Usable at 200% zoom. Touch targets meet 44x44 minimum.

How It Works

From Endpoint to Alert

1

Register

Add endpoints to monitor with URL, expected status, check interval, and timeout thresholds.

2

Poll

APScheduler fires checks on schedule. Records status code, response time, and content hash for each endpoint.

3

Detect

Failed checks trigger incidents automatically. Degraded checks fire when latency exceeds thresholds. WebSocket broadcasts the change instantly.

4

Resolve

When the endpoint recovers, the incident auto-resolves. Uptime stats recalculate. The alarm banner clears.

Technical Specifications

Under the Hood

Backend

  • FastAPI (Python 3.12+)
  • PostgreSQL + SQLAlchemy 2.0 async
  • APScheduler for check scheduling
  • Alembic migrations
  • avian-diagnostics integration

Frontend

  • React 19 + TypeScript
  • Vite build system
  • Recharts for latency charts
  • Framer Motion animations
  • CSS Modules with AVIAN design system
  • Light and dark mode

Real-Time

  • WebSocket for live updates
  • Check result broadcasting
  • Incident state changes
  • No polling required

Monitoring

  • HTTP status validation
  • Response time measurement
  • Content hash verification
  • Configurable check intervals
  • Automatic incident lifecycle

Development

100% Built by Claude

Vigil was built entirely by Claude (Anthropic) working alongside a single human supervisor. Every line of code, every test, every deployment: AI-authored with human direction.