research-portal 0.1.4

Zero-config research workstation dashboard with automatic pipeline discovery

research-portal

Zero-config research workstation dashboard with automatic pipeline discovery.

Drop it on any Linux research machine and instantly get live GPU/CPU monitoring, running pipeline detection, and a resource map -- no configuration needed.

Installation

pip install research-portal

Or install from source:

git clone https://github.com/ahb-sjsu/atlas-portal.git
cd atlas-portal
pip install -e ".[dev]"

Quick Start

research-portal

That's it. Open the URL printed in the terminal (default: http://0.0.0.0:8080).

Default credentials: atlas / atlas2026!research (override with env vars or CLI flags).

Screenshots

Dashboard, Resource Map, and Pipeline Flow views coming soon.

Features

• Auto-detects hardware -- hostname, CPU model, GPU models, RAM, disk (reads from /proc, nvidia-smi, etc.)
• Live system monitoring -- CPU temperatures, GPU utilization/memory/thermals, RAM, disk, load averages
• Pipeline discovery -- automatically finds running Python/bash pipelines from process list
• Resource map -- per-core CPU utilization grid with job labels, per-GPU visual bars
• Pipeline flow -- live view of all discovered pipelines with stage progress
• Security -- HTTP Basic auth, security headers (CSP, X-Frame-Options, etc.)
• SSL support -- auto-detects cert.pem/key.pem for HTTPS
• Zero config -- works on any Linux workstation out of the box

Pages

Route	Description
/	Main dashboard -- temperatures, GPU status, memory, active sessions
/map	Resource map -- per-core CPU grid, per-GPU utilization visuals
/flow	Pipeline flow -- auto-discovered running/completed pipelines

API Endpoints

Endpoint	Description
GET /api/status	System metrics (CPU temps, GPUs, memory, disk, load, sessions)
GET /api/cores	Per-core utilization with job labels
GET /api/pipelines	Auto-discovered pipeline list with stages
GET /api/system-info	Static hardware info (hostname, CPU, GPU models, RAM)

Configuration

CLI Flags

research-portal [OPTIONS]

  --host HOST       Bind address (default: 0.0.0.0)
  --port PORT       Port (default: 8443 with SSL, 8080 without)
  --no-auth         Disable HTTP Basic authentication
  --user USER       Override auth username
  --password PASS   Override auth password
  --no-ssl          Force plain HTTP even if certs are present
  --version         Print version and exit

Environment Variables

Variable	Description	Default
PORTAL_USER	Auth username	atlas
PORTAL_PASS	Auth password	atlas2026!research
PORTAL_SECRET	Flask secret key	random
PORTAL_RESULTS_DIR	Directory for downloadable result files	.

SSL

Place cert.pem and key.pem in the working directory or package directory. The portal will auto-detect them and serve on HTTPS port 8443.

Generate self-signed certs for development:

openssl req -x509 -newkey rsa:4096 -keyout key.pem -out cert.pem -days 365 -nodes

Development

pip install -e ".[dev]"
pytest -v
ruff check src/ tests/
ruff format src/ tests/

License

MIT