Skip to content
Tech wire
North Korea’s hackers using AI for attacks, cybersecurity firm saysAl JazeeraIndustry Leaders Unite in Open Secure AI Alliance for AI Safety and SecurityNVIDIA BlogWhat Is Open-Weights A.I.?The New York TimesAttackers are targeting open-source AI just as Big Tech is embracing itDefense OneOpen Weights and American AI LeadershipMicrosoftHugging Face hack marks start of dangerous AI cyber era and many firms 'don't even know it'CNBCMicrosoft, tech companies throw weight behind spread of open-source AICyberScoopAsserting American Leadership in Open Source AIAndreessen HorowitzCISA guidance targets water sector security, open source AI and moreFederal News NetworkTech giants launch open-source AI security alliance after attackTech XploreCompanies rally behind open-source techThe HillAI & Tech Brief: The open-source debateThe Washington PostAlibaba targets Nvidia’s dominant software ecosystem with open-source AI stackSouth China Morning PostChina's 'open' AI is a terrible business, and nothing like open-source softwareBusiness Insider

karpathy/autoresearch

Saturatedrank #92

AI agents running research on single-GPU nanochat training automatically

93.5k+2.1k · 7d 13.3k 442est. NOASSERTIONPythonpushed 5mo ago
Compare View on GitHub

Composite

58

conf 65%

Momentum

65

conf 60%

Vitality (PVI)

50

conf 67%

Community

73

conf 60%

Security & maint.

25

conf 85%

Ecosystem influence

76

conf 60%

Opportunity

67

conf 60%

Breakout · 7d

13%

conf 71%

Executive summary

autoresearch is currently classified as saturated with a composite score of 58/100. Momentum (65) and vitality (50) diverge — attention is running ahead of maintenance depth. Estimated 7-day gain of 2,076 stars against a 93.5k-star base. Risk is assessed as high; the interpretable model puts its probability of entering the top trend cohort within 7 days at 13%.

90-day trajectory

+15.2k · 30d
2026-05-132026-08-10

Score breakdown

Every input, weight and normalized value — nothing hidden. Inputs marked est. are estimated until the full ingestion pipeline lands.

Momentum

Star velocityest.25%92
Star accelerationest.15%29
Fork velocityest.10%91
Contributor growthest.10%100
Issue activity growthest.10%49
PR activity growthest.10%76
External mention velocityest.10%20
Release event impactest.5%40
Trending-rank persistenceest.5%36

Model tn-interpretable-1.0.0 · confidence 60% · coverage 100%

Project Vitality Index

Commit consistency20%0
Release cadenceest.15%41
PR responsivenessest.15%53
Issue responsivenessest.15%52
Contributor diversityest.10%98
Maintainer continuityest.10%58
Documentation qualityest.5%83
Dependency freshnessest.5%88
Security hygieneest.5%70

Model tn-interpretable-1.0.0 · confidence 67% · coverage 92% · missing: licenseClarity

Security & maintenance

Release recency25%0
License clarity20%30
Maintenance status25%0
Dependency stalenessest.15%40
Bus factorest.15%86

Model tn-interpretable-1.0.0 · confidence 85% · coverage 100%

Ecosystem influence

Dependent repositoriesest.30%99
Quality-weighted dependentsest.20%84
Documentation citationsest.15%68
Contributor overlapest.15%76
Registry adoptionest.10%70
Topic centralityest.10%5

Model tn-interpretable-1.0.0 · confidence 60% · coverage 92% · missing: packageRegistryAdoption