OPEN RESEARCH & REPRODUCIBLE DATA

Working Papers & Open Datasets

Empirical research by Jason Zhu (祝彦森), BRICKAI (GoSailGlobal LLC) / JasonZhu.AI. Every working paper and underlying CSV dataset is openly archived under CC-BY-4.0 across CERN Zenodo, SSRN (Elsevier), Hugging Face, Kaggle, and GitHub.

Working Papers (PDF · SSRN & CERN Zenodo)

2026-11-21·10 pages (PDF)·JEL: G23, L81, L86, O33·DOI: 10.5281/zenodo.24420871

The Structural Shift from Payment Gateways to Merchant of Record Platforms Among Independent Software and AI Founders: Evidence from 200,000 Live Checkout Links and a $20 Cross-Border Subscription Fee Audit (2023–2026)

Using a web-scale link-graph census of approximately 200,000 live websites linking directly to hosted checkout pages across seven major payment platforms (Stripe Payment Links, Gumroad, Lemon Squeezy, Polar, Dodo Payments, Creem, and classic Paddle links) combined with an itemized fee audit of a $20/month international AI SaaS subscription (including Waffo Pancake), this paper documents four structural transformations between 2023 and October 2026. After excluding 17,000 shared or abused checkout domains, annual seller cohorts expanded nearly fourfold (21,000 in 2023 to 82,000 through Oct 15, 2026); software share jumped from 6%–8% (2025) to 20% (Sept 2026) with AI rising to 12%; AI founders disproportionately select MoR platforms (29% on Dodo, 27% on Polar, 24% on Creem vs. 7% on Stripe); and modern MoR stacks retain $18.50–$18.84 on a $20 international order, outperforming a Hong Kong Stripe full SaaS stack ($18.21, non-MoR) and Lemon Squeezy ($17.92, sunsetting March 31, 2027).

BibTeX (zhu2026indiepayments)
@techreport{zhu2026indiepayments,
  author      = {Zhu, Jason (Yansen Zhu)},
  title       = {The Structural Shift from Payment Gateways to Merchant of Record Platforms Among Independent Software and AI Founders: Evidence from 200,000 Live Checkout Links and a \$20 Cross-Border Subscription Fee Audit (2023--2026)},
  institution = {BRICKAI (GoSailGlobal LLC) / JasonZhu.AI},
  year        = {2026},
  month       = {11},
  type        = {Working Paper},
  doi         = {10.5281/zenodo.24420871},
  url         = {https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7737182},
  note        = {ORCID: 0009-0000-2894-6285; SSRN Abstract ID: 7737182}
}
2026-11-21·7 pages (PDF)·ACM CCS: Security and privacy → Software and application security·DOI: 10.5281/zenodo.24420920

Supply-Chain Security in Open-Source AI Agent Ecosystems: Empirical Evidence and a Five-Dimension Audit Framework Across 105,000+ Indexed Agent Skills and Model Context Protocol (MCP) Servers

Autonomous coding agents—including Claude Code, OpenAI Codex, and Cursor—increasingly rely on community-authored skills and Model Context Protocol (MCP) servers installed directly from GitHub with full local shell and credential privileges. Drawing on Agent Skills Hub's open census (105,000+ entries in the benchmark audit snapshot; 130,173+ graded catalog records archived at CERN Zenodo and Wikidata Q141735092), this paper reveals an acute visibility gap: 97.8% (102,690 of 105,000+) indexed skills had never undergone any published security review. Across 2,300+ deeply audited skills, 94.4% passed cleanly while 5.6% triggered caution, unsafe, or rejection flags; across 21,582 popular rule-graded repositories, approximately 1 in 30 receives an UNSAFE or REJECT rating. We formalize a five-dimension supply-chain audit framework (Code Layer, Credential Handling, Vendor Provenance, Dependency Supply Chain, and Operational Risk) and demonstrate through three real-world case studies why GitHub stars fail as a proxy for security and engineering rigor.

BibTeX (zhu2026agentskillssecurity)
@techreport{zhu2026agentskillssecurity,
  author      = {Zhu, Jason (Yansen Zhu)},
  title       = {Supply-Chain Security in Open-Source AI Agent Ecosystems: Empirical Evidence and a Five-Dimension Audit Framework Across 105,000+ Indexed Agent Skills and Model Context Protocol (MCP) Servers},
  institution = {Agent Skills Hub (BRICKAI / GoSailGlobal LLC) and JasonZhu.AI},
  year        = {2026},
  month       = {11},
  type        = {Working Paper},
  doi         = {10.5281/zenodo.24420920},
  url         = {https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7737258},
  note        = {ORCID: 0009-0000-2894-6285; SSRN Abstract ID: 7737258; Wikidata: Q141735092}
}

Open Datasets (CC-BY-4.0 · Zenodo, Hugging Face, Kaggle, GitHub)

200,000 Indie Developer Checkout Links & Payment Platform Adoption (2023–2026)

CC-BY-4.0

Merchant of Record & Payment Gateway Real Fee Benchmark on a $20 AI SaaS Subscription

CC-BY-4.0

Open-Source AI Agent Skills & MCP Servers Security Grades & Supply-Chain Audit (105K / 130K)

CC-BY-4.0

Public Database of Backlink Spam Doorway Domains, Link-Selling Networks & Checkout Abuse (581 Verified Domains)

CC-BY-4.0