DevPulse

Real-time, local-first system observability for developers.
Monitor CPU, memory, disk, network, processes, alerts, and historical metrics from one responsive dashboard.

Why DevPulse?

Cloud monitoring products are excessive when you only need a fast answer to what is happening on this machine? DevPulse keeps collection and storage local, requires no account, and exposes both a polished dashboard and an OpenAPI-documented API.

Features

• Live WebSocket telemetry for CPU, memory, disk, network, process count, and uptime
• Process explorer ranked by resident memory usage
• Threshold alerts configurable through environment variables
• SQLite history with bounded API queries
• Interactive API docs generated by FastAPI
• Automatic reconnect when the WebSocket connection drops
• Responsive dashboard with zero frontend dependencies
• Cross-platform collection through psutil (Windows, Linux, macOS)
• Docker support, health check, non-root container, CI, lint, tests, and coverage gate

Preview

The dashboard is designed as a developer observability console: live metric cards, animated sparklines, transfer visualization, threshold warnings, and a real-time process table.

Quick start

Requirements

• Python 3.11+
• uv (recommended)

Run locally

git clone https://github.com/andraokta/devpulse.git
cd devpulse
uv sync --extra dev
uv run devpulse

Open:

• Dashboard: http://127.0.0.1:8000
• API documentation: http://127.0.0.1:8000/docs
• Health endpoint: http://127.0.0.1:8000/api/health

Run with Docker

docker compose up --build

Configuration

Copy .env.example to .env and override what you need:

Variable	Default	Purpose
DEVPULSE_HOST	127.0.0.1	Bind address
DEVPULSE_PORT	8000	HTTP port
DEVPULSE_SAMPLE_INTERVAL	2	Seconds between samples
DEVPULSE_DATABASE_PATH	data/devpulse.db	SQLite location
DEVPULSE_CPU_THRESHOLD	85	CPU alert percentage
DEVPULSE_MEMORY_THRESHOLD	90	Memory alert percentage
DEVPULSE_DISK_THRESHOLD	90	Disk alert percentage

API

Method	Endpoint	Description
GET	/api/health	Service health and version
GET	/api/metrics/current	Current metrics and active alerts
GET	/api/metrics/history?limit=60	Historical samples
GET	/api/processes?limit=10	Top processes by memory
WS	/ws/metrics	Real-time metric stream

Example:

curl http://127.0.0.1:8000/api/metrics/current

Development

uv sync --extra dev
uv run ruff check .
uv run pytest --cov=devpulse --cov-report=term-missing
uv build

Tests cover metric normalization, process ordering, alert evaluation, API health, current metrics, process limits, and invalid request validation.

Architecture

Operating System ──psutil──> Collector ──> Alert Engine
                                 │
                                 ├──────> SQLite history
                                 ├──────> REST API / OpenAPI
                                 └──────> WebSocket ──> Dashboard

See docs/ARCHITECTURE.md for component boundaries, data flow, and security considerations.

Security

DevPulse binds to localhost by default. Process names and system metrics may be sensitive; if you expose the service to a network, place it behind an authenticated reverse proxy.

• No shell execution
• Parameterized SQLite queries
• Bounded query parameters
• Non-root Docker runtime
• Read-only GitHub Actions permissions

Roadmap

• Per-process CPU trend charts
• Export history to CSV
• Configurable retention policy
• Optional authenticated remote mode
• Prometheus-compatible endpoint

Contributing

Issues and pull requests are welcome. Please run lint and tests before submitting changes.

License

MIT © 2026 Andra Dwi Okta Ramadhan