Crucix vs worldmonitor
Self-hosted OSINT terminal: 27 open feeds — satellite fires, flights, radiation, sanctions, markets, conflict data — polled in parallel onto one Jarvis-style dashboard. LLM turns it two-way. — versus — Real-time global situational-awareness dashboard: curated feeds AI-synthesized into briefs over a 3D globe, cross-stream correlation and a country instability index — plus MCP, REST API and CLI.
The same job — fuse dozens of open feeds into one live situational board. Crucix is a Jarvis-style terminal over 27 feeds you self-host; World Monitor is a mapped web/desktop app with themed variants, instability scoring and an agent-facing API. Terminal minimalism versus mapped breadth.
| Crucix | worldmonitor | |
|---|---|---|
| Stars | 12k | 85k |
| Forks | 1.8k | 13k |
| Language | JavaScript | TypeScript |
| License | AGPL-3.0 | AGPL-3.0 |
| Last activity | 3 months ago | today |
| Topics | security, web | security, web |
| Curated connections | 3 | 3 |
Crucix — the curator's take
The fun kind of serious: one Node process (a single express dependency!) pulls 27 public intelligence feeds every 15 minutes onto a self-contained dashboard with a 3D globe — and hooking up an LLM makes it two-way, pushing tiered alerts to Telegram/Discord and answering /brief from your phone. Zero cloud, genuinely yours. NOT an investigation platform: it monitors the world's pulse, it doesn't build case files on entities (that's flowsint's job) — and treat the feed quality honestly: public OSINT sources carry noise, coverage gaps and occasional propaganda; alerts are leads, not conclusions. AGPL if you're embedding it commercially.
worldmonitor — the curator's take
An OSINT wall that agents can also read. Feeds across geopolitics, finance, energy, climate, aviation, cyber and military land on a globe.gl/deck.gl map with cross-stream correlation and a server-authoritative Country Instability Index, and the whole thing is one codebase producing six themed variants plus a Tauri desktop binary. What makes it catalog-worthy rather than a nice dashboard: an MCP server, OpenAPI REST base, npm CLI, Python/Ruby/Go SDKs, llms.txt and an agent-skills manifest — it is designed to be consumed by an agent, not just watched. Local AI via Ollama means it can run with no keys at all. Caveats: AGPL-3.0-only, so commercial SaaS use carries copyleft obligations; several feeds need your own credentials; and the hosted API key gates the interesting `tools/call` paths.