OSSINTEL
New — manipulation detection v2 with IP/ASN clustering

Trust intelligence for open source.

Bloomberg Terminal for GitHub. Score any repository for trust, manipulation risk, organic growth, security hygiene, and ecosystem adoption — in seconds.

Try:langflow-ai/langflowvercel/next.jsggerganov/llama.cpp
TRENDING SUSPICIOUS
View all
TOP ORGANIC GROWTH
View all
REALTIME FEED
LIVE
  • 12:04:22Star burst · langflow-ai/langflow — +1,240 stars in 14 minHIGH
  • 12:03:11Release v2.41 · supabase/supabase — Tag pushedLOW
  • 12:02:48Lockstep stars · openagents-corp/swarmflow — 94% accounts < 30dHIGH
  • 12:01:33PR merged · vercel/next.js — Stable channelLOW
  • 12:00:51IP cluster · rapid-labs/rapid-llm — 47 stars / 3 ASNHIGH
  • 11:59:02HN front page · ggerganov/llama.cpp — Organic surgeLOW
  • 11:58:14Discord mention · shadcn-ui/ui — +212 referralsLOW
  • 11:57:39Suspicious fork · Significant-Gravitas/AutoGPT — Empty fork patternMEDIUM
CORE THESIS

Why stars are broken.

Stars used to be a proxy for usefulness. Today they are a market — bought, farmed, and traded. OSSINTEL separates real adoption from manufactured momentum so investors, security teams, and developers can trust what they see.

Lockstep starring
Hundreds of accounts star within the same 90-second window from clustered IP ranges.
Burst-only growth
Stars arrive in step-functions tied to bot waves, not sustained organic interest.
Account aging fraud
47% of stargazers are < 30 days old. Ecosystem average is 6%.
Broken trust signal
Star count alone correlates ~0.31 with downstream adoption — barely better than chance.
ORGANIC vs MANIPULATED — typical pattern
Organic
Manipulated
USE CASES

Used by investors, security teams, and developers.

Investors
Diligence on OSS-native investments. Detect manufactured momentum before term sheets.
  • Pre-seed → growth screening
  • Cohort comparisons
  • Manipulation evidence packs
Security teams
Track risky dependencies, suspicious maintainer activity, and supply chain anomalies.
  • Maintainer reputation
  • Dependency trust scoring
  • Anomaly alerts to Slack/PagerDuty
Developers
Pick libraries you can rely on. Stop being burned by repos that look popular but aren't.
  • Browser overlay on github.com
  • Compare alternatives
  • Watchlist your stack
LEADERBOARD

Trust ranking — last 7 days

Open watchlist →
#RepositoryTopicStarsΔ 7dRiskTrustTrend
1vercel/next.js
The React framework for the web
Framework124.3K1.8%LOW96
2supabase/supabase
The open source Firebase alternative
Database79.1K1.1%LOW93
3ggerganov/llama.cpp
Port of Facebook's LLaMA model in C/C++
Inference71.2K2.3%LOW91
4shadcn-ui/ui
Beautifully designed components built with Radix UI
UI76.9K3.2%LOW89
5langflow-ai/langflow
Low-code LLM app builder
AI/ML147.6K4.2%HIGH61
6Significant-Gravitas/AutoGPT
Autonomous AI agents framework
Agents168.2K7.1%HIGH54
7rapid-labs/rapid-llm
Fast inference toolkit
Inference9.2K9.1%HIGH42
8openagents-corp/swarmflow
Multi-agent orchestration runtime
Agents18.4K12.4%HIGH38
SCORING MODEL

A weighted, auditable trust score.

Every score is reproducible. We surface the underlying signals, weight them transparently, and let your team override the model with policy.

  • • Per-signal contribution attribution
  • • Counterfactual “what-if” simulations
  • • Policy overlays (e.g. block repos < trust 70 in production)
  • • Evidence-backed alerts with raw data export
SIGNAL WEIGHTS
Repository Analysis
Drill into trust score, manipulation evidence, and adoption signals.
Watchlist Dashboard
Track your portfolio, dependencies, and competitor repos in one screen.
Browser Overlay
See OSSINTEL trust scores directly on github.com.
© 2026 OSSINTEL · Trust intelligence for open source · Demo build