Research Atlas
Curated map of the repo's papers, synthesis docs, and hidden implementation context.
The repo contains far more research than the old docs navigation suggested. This page surfaces the highest-value source material so architecture, design, and product work can start from real evidence instead of rediscovering it from scratch.
The Four Documents To Read First
| Document | Why it matters |
|---|---|
docs/research/RESEARCH_INVENTORY.md | Complete inventory of papers, math extractions, and visualization references |
docs/research/analysis/MASTER-SYNTHESIS.md | Consolidated analysis across 80+ papers and theme clusters |
docs/research/analysis/APPLICATION-STATE-OF-PLAY.md | Gap analysis between theory and what the code actually implemented at audit time |
docs/architecture/unified-architecture-state.md | Latest implementation tracker that supersedes older state snapshots where they disagree |
What Is Hidden Gold
Research And Math
- Multilayer network theory, multiplex formalism, and participation-coefficient ideas
- FX topology, MST/PMFG, contagion, routing, and arbitrage research
- Token-economics math ranging from CES and HHI through entropy, Gini, and emission models
- Reputation, governance, conviction voting, Sybil resistance, and identity papers
- Visualization research for stacked layers, multilevel graphs, and financial dashboards
Design And Brand
docs/design/BRAND_FINALIZATION_2026-03-03.mdlocks the UI color semanticsdocs/design/BRAND_COLOR_SEMANTICS.mdexplains which colors are semantic UI tokens versus visualization-only FX colorsdocs/research/fxyz-brand-codex-final.jsxcarries the detailed brand codex source work
Implementation Archaeology
archive/work-logs/records deployment fixes, Fumadocs migration history, docs cleanup, and production issue contextcontext/contains extracted notebook material, scanned-source summaries, and structured idea inventories that now carry a lot of the project's hidden source-of-truth context.claude/contains operational memory, agents, cache material, and other machine-generated context that can explain how certain decisions emerged
Extracted Notebook Corpus
Some of the strongest conceptual material is now in context/, not only in docs/research/.
context/notebook-17-fxyz-core-network-vision.mdcaptures the core business-model loop, network projection theory, and early f(x,y,z) framingcontext/notebook-26-network-architecture.mdcontains corridor architecture, identity stack, and micro-transaction system-design notescontext/notebook-27-token-transaction-architecture.mdmaps transaction flows, entity layers, and early token transfer abstractions
Treat these as high-value conceptual archaeology. They are not polished docs, but they contain real design intent.
How To Use The Research Corpus
If you are shaping the product
- Start with the research inventory and master synthesis.
- Pull only the ideas that map to current product surfaces and architecture constraints.
If you are updating the docs
- Use the unified architecture tracker plus live code for "what is true now."
- Use research docs to explain why the system is shaped that way.
If you are planning engineering work
- Mine the "implementation roadmap" sections in the synthesis docs.
- Cross-check against the latest current-state tracker before treating any roadmap item as still open.
Older research snapshots are still valuable, but some of them deliberately document gaps that were later closed. The best example is contribution scoring: February audit notes describe a missing CES implementation, while the March architecture tracker and current code show that alignment work was completed.