27 Intelligence Feeds, One Dashboard, Zero Cloud: How Crucix Became GitHub’s Hottest OSINT Project

Imagine getting a Telegram alert at 2 AM because radiation levels near a nuclear site just spiked — cross-referenced with unusual military flight activity in the same region and a surge in conflict-related social media chatter. That is the kind of scenario Crucix was built for, and it is why the open-source community cannot stop talking about it.

Crucix is a self-hosted personal intelligence agent that pulls data from 27 public sources — satellite fire detection, flight tracking, radiation monitoring, economic indicators, conflict data, sanctions lists, and more — every 15 minutes, rendering it all on a single Jarvis-style dashboard running on your own machine. Hook up an LLM, and it becomes a two-way assistant you can query from your phone.

The project hit GitHub Trending in mid-March 2026, racked up over 2,700 stars, and landed on trendshift.io as a top AI-related repository. The OSINT community noticed.

What Crucix Actually Does

At its core, Crucix runs a sweep cycle every 15 minutes. All 27 data sources fire simultaneously via Promise.allSettled(), meaning a single failed API does not crash the entire pipeline. The system computes deltas — what changed, what escalated, what de-escalated — and classifies signals into three alert tiers:

FLASH: Critical cross-domain signals demanding immediate attention
PRIORITY: Significant changes or escalations worth monitoring
ROUTINE: Baseline informational updates

These alerts push to Telegram and Discord with color-coded formatting. The Telegram bot responds to commands like /brief for a compact intelligence summary, /sweep to trigger a manual data pull, and /status for system health checks.

The dashboard itself is a self-contained HTML file powered by a 3D WebGL globe (Globe.gl) with nine marker categories — fires, aircraft, radiation sites, maritime activity, SDR signals, OSINT events, health alerts, news hotspots, and conflict zones. Six region filters let you rotate and zoom to specific areas. There are live market tickers from Yahoo Finance, risk gauges for VIX and credit spreads, and a news feed merging RSS, GDELT headlines, and 17 curated Telegram intelligence channels.

The 27 Data Sources, Broken Down

The scope of what Crucix aggregates is what sets it apart from typical OSINT dashboards. The sources span five tiers:

Geopolitical and OSINT (11 sources): GDELT for global event tracking across 100+ languages, OpenSky and ADS-B Exchange for real-time flight tracking, NASA FIRMS for satellite fire and thermal detection, Maritime AIS for vessel tracking and sanctions evasion detection, Safecast for citizen-science radiation monitoring near nuclear sites, ACLED for armed conflict events, ReliefWeb for UN humanitarian data, WHO for disease outbreak alerts, and both OFAC and OpenSanctions for sanctions screening.

Economic and Financial (7 sources): FRED covers 22 economic indicators including yield curves and CPI. US Treasury provides national debt and yield data. BLS handles employment figures. EIA tracks energy prices. The NY Fed’s Global Supply Chain Pressure Index, USAspending for defense contracts, and UN Comtrade for strategic commodity trade flows round out the financial picture.

Environment, Social, and SIGINT (7 sources): NOAA weather alerts, EPA RadNet radiation monitoring, USPTO patent filings in strategic tech areas, Bluesky and Reddit sentiment analysis, Telegram channel scraping, and KiwiSDR — a global network of roughly 600 HF radio receivers.

Space (1 source): CelesTrak for satellite tracking, recent launches, ISS position, military constellations, and Starlink/OneWeb fleet counts.

Markets (1 source): Yahoo Finance for real-time prices across indices, crypto, commodities, and energy.

The important detail: 18 of these sources work without any API keys at all. The free-tier keys (FRED, NASA FIRMS, EIA) take minutes to obtain. Only ADS-B Exchange requires a paid subscription (~$10/month via RapidAPI) for unfiltered military tracking.

LLM Integration: From Dashboard to Analyst

Without an LLM, Crucix is a data aggregator. With one, it becomes something closer to an intelligence analyst. The system supports four providers — Anthropic Claude, OpenAI, Google Gemini, and OpenAI Codex — all implemented via raw fetch calls with no SDK dependencies.

When an LLM is connected, two features unlock. First, AI trade ideas: the system generates 5-8 quantitative analyst-style suggestions citing specific data points from the latest sweep. Second, smart alert evaluation: instead of rule-based classification, the LLM evaluates signals with cross-domain correlation and confidence scoring. A radiation spike alone might be ROUTINE; a radiation spike plus unusual military flights plus a sanctions list update in the same region gets flagged as FLASH.

If the LLM provider goes down, Crucix falls back to rule-based evaluation automatically. LLM failures never crash sweep cycles — a design choice that reflects the project’s philosophy of graceful degradation.

How Crucix Stacks Up Against Alternatives

Crucix enters a growing field of open-source OSINT dashboards, but its positioning is distinct.

World Monitor is probably the closest competitor — another open-source real-time intelligence dashboard that consolidates news and geopolitical signals into a map-based interface. It is more focused on news aggregation and geopolitical monitoring, whereas Crucix leans harder into cross-domain data fusion (combining financial, environmental, space, and SIGINT data alongside geopolitical feeds).

Monitor The Situation offers a more polished UI and is often praised for design and usability, but it is not open-source and lacks the self-hosted, zero-telemetry angle that drives much of Crucix’s appeal.

Taranis AI is an enterprise-grade OSINT tool using AI for information gathering and situational analysis, built on a Celery-based async task queue. It is more mature but also more complex to deploy and oriented toward organizational use rather than individual analysts.

On the commercial side, platforms like Flashpoint and Dataminr can run into six-figure annual contracts. Crucix’s pitch is straightforward: comparable cross-domain awareness for the cost of a VPS and a few free API keys.

The tradeoff is real, though. Enterprise tools come with curated data, dedicated support, and validated intelligence pipelines. Crucix relies on public APIs that can change or rate-limit without notice, and the quality of LLM-generated analysis depends entirely on your chosen provider and prompt engineering.

What the Community Is Saying

The Hacker News discussion around Crucix (and similar OSINT dashboards) reveals a split reaction.

On the positive side, developers praised specific technical choices. One commenter called the live GPS jamming calculation using commercial flight NAC-P degradation “honestly brilliant.” Others appreciated the Jarvis aesthetic — “nothing wrong with that, beats a boring corporate dashboard.” Security audits by community members found no malicious behavior in the codebase.

The criticism is equally pointed. Multiple users reported issues getting the project running, citing missing data and broken dependencies. Some questioned whether the codebase was partially AI-generated, noting placeholder data in certain modules. One user flagged that “UAV data is fake.” And there is a philosophical objection worth considering: does constant awareness of global crises you cannot influence create “hyper-alert stress” rather than actionable intelligence?

The project also has a fork (spiderdev27/Crucix-intelligence-terminal) maintaining a parallel implementation, which adds to the sense of active community interest — though it also raises questions about which version to follow.

Running It Yourself

Setup is minimal by design. The project runs on Node.js 22+ with Express as the only required dependency:

git clone → npm install → cp .env.example .env → npm run dev

The dashboard launches at localhost:3117. Docker Compose is also supported for containerized deployment with persistent data. A diagnostic script (node diag.mjs) checks Node version, module imports, and port availability if something goes wrong.

The architecture is pure ESM — all .mjs files with explicit imports. Each of the 27 data sources lives in its own standalone file under apis/sources/, meaning you can test any single source independently with node apis/sources/[name].mjs. The AGPL v3 license ensures the project stays open-source.

FAQ

Is Crucix free to use?
Yes. Crucix is fully open-source under AGPL v3 with no subscription or cloud service. Most data sources work without API keys, and the ones that need keys offer free tiers. The only potentially paid source is ADS-B Exchange at roughly $10/month for unfiltered flight tracking.

What are the main competitors to Crucix?
World Monitor (open-source, news-focused), Monitor The Situation (polished UI, closed-source), and Taranis AI (enterprise-grade, open-source) are the closest alternatives. Commercial options like Flashpoint and Dataminr offer similar capabilities but at enterprise pricing.

Who is Crucix designed for?
Researchers, journalists, traders, OSINT analysts, and security professionals who want cross-domain situational awareness without relying on expensive enterprise platforms. The self-hosted, zero-telemetry design also appeals to privacy-conscious users.

Do I need an LLM to use Crucix?
No. The dashboard and all 27 data sources work without any LLM provider. Adding an LLM (Anthropic, OpenAI, Gemini, or Codex) unlocks AI-powered trade ideas and smarter alert classification, but the core functionality is fully independent.

How reliable is the data?
Crucix aggregates from established public sources (NASA, NOAA, WHO, FRED, etc.), so data quality matches those upstream providers. However, public APIs can rate-limit or change without notice, and the system’s graceful degradation means a failed source simply shows an error rather than crashing. The LLM analysis layer adds interpretation that should be treated as supplementary, not authoritative.

Top AI Product

Leave a comment Cancel reply