Agent operations
Multi-agent execution patterns for long research, coding support, project handoff, system diagnostics, and reliability escalation.
AI LAB · PUBLIC CONTENT LAYER
Experiments that turn personal AI into operating systems
Solarfans AI Lab is the public index for Melvin's agent workflows, knowledge infrastructure, automation hooks, trading research, and edge-service experiments. The public layer shows intent and architecture while private controls stay offline or behind protected channels.
LAB TRACKS
Four working tracks behind the AI Lab.
Each track is designed to be useful publicly without leaking operational secrets or turning the site into a control panel.
Knowledge workflows
Capture, distillation, Obsidian publishing, Hindsight recall, skill routing, and reset-safe project continuity.
Automation hooks
Signed webhooks, watchdogs, scheduled reports, and event-driven workflows that turn external signals into controlled tasks.
Research systems
Trading research, AI infrastructure scouting, web intelligence, and reusable analysis pipelines for decision support.
DESIGN PRINCIPLE
Public by default for narrative, private by default for control.
The lab page explains what is being built. Operational dashboards, credentials, model quotas, tokens, SSH details, and VPN configuration stay out of the public web surface.
Publicproject narrative, status summaries, selected notes
Protectedagent dashboards, logs, quota details, admin tools
Privatesecrets, tokens, VPN URI, SSH and tunnel credentials