Map of Augments
Semantic explorer for traversing the OPTX documentation graph — cognitive cartography, AGT-classified.
Map of Augments
A navigable knowledge map of every page in the OPTX documentation. Each node is classified by its primary AGT tensor — the cognitive dimension it engages. Hover to reveal connections. Filter by tensor or search to isolate clusters.
Getting Started
3Dual-mode JettChat overview, naming hierarchy, protocol primitives
Two JettChat modes (xChat Native + Phantom Mode) sharing AGT, JTX, Solana
Solana program addresses, token mints, wallets
JettChat
4Encrypted AI chat — two modes (xChat Native + Phantom Mode)
X OAuth 2.0 PKCE + Solana JTX gating + Ed25519 JWT
Tor + post-quantum (X25519 + ML-KEM-1024) + StrongBox/TEE + duress PIN
E2EE messaging shared by both modes — gaze cursor, offline-first, groups, self-destruct
Authentication
3Unified auth surface — supports both xChat Native and Phantom Mode
AGT biometric auth — iris tracking, tensor classification
ERC-8004 soulbound wallet for computational identity (roadmap)
Token
3JettChat access + governance token on Solana mainnet
Canonical 3-tier model (MOJO / DOJO / SPACE COWBOY) — stake or subscribe
Wallet-less paths — Stripe + Tempo CLI
Protocol — AARON
5Biometric proof protocol and edge router
Gaze-derived cryptographic proofs
AARON verification flow step-by-step
Integrating AARON into client applications
Internal architecture of the AARON router
AstroJOE — Agent OS
8Intelligent agent orchestrating the OPTX ecosystem
SKILL.md format, progressive disclosure, augments
SpacetimeDB memory — tables, reducers, API
Task lifecycle, DAG swarm coordination
Multi-API AI gateway on edge hardware
Hermes OPTX API endpoints and configuration
The Curator release — autonomous skill curator, +4 providers, ComfyUI + TouchDesigner-MCP bundled, ~57% TUI cold-start cut
Optional federation transport for AstroJOE — not part of JettChat dual-mode core
Architecture Flows
8Mermaid diagrams for every major system flow
Happy path: task creation to completion
Multi-agent DAG decomposition
Gaze-gated task authorization flow
XRPL → Wormhole → Solana pipeline
Identity sources → on-chain registration
State machine: Discovered → Completed/Failed
Full network map of all OPTX services
Infrastructure
2On-Chain Bridge
2Solana-native bridge hub — EVM and XRPL bridging on roadmap
Home chain — $OPTX, $JTX, $CSTB, Metaplex identity
Reference
4DOJO
3How to Read This
The MOA is organized by conceptual domain. Each section is a cluster in the documentation graph, and each entry is a traversable node. The small number on the right is the edge count — how many other pages this node connects to. Denser nodes are more central to the system.
Hover any node to see its immediate connections light up across all sections. The links row shows which pages reference each other — these are the edges of the knowledge graph, rendered as traversable text instead of drawn lines.
AGT — Augmentive Gaze Tensor
The classification system behind this graph evolved through three generations:
| Generation | Name | Focus |
|---|---|---|
| v1 | Adaptive Gaze Tensor | Reactive biometric classification |
| v2 | Agentive Gaze Tensor | Agent-driven tensor inference |
| v3 | Augmentive Gaze Tensor | SDK for agentic Map of Context |
AGT v3 is the current system — an SDK that lets agents classify, navigate, and augment any knowledge graph using the three-tensor model. Every page in OPTX carries one of three tensor tags:
- COG — Cognitive. Analytical pages that require sustained reasoning: architecture, protocol internals, reference material.
- EMO — Emotional. Relational pages about identity, personality, agent interaction, human-facing systems.
- ENV — Environmental. Spatial pages mapping networks, infrastructure, addresses, physical topology.
Use the tensor filter in the sidebar to isolate one dimension and see which pages belong to it. The distribution reveals the shape of the system — ENV-heavy means infrastructure-dense, COG-heavy means architecturally complex, EMO-heavy means agent-facing.
Jett Cursor
The Jett Cursor is an AGT simplex widget — a triangular indicator that visualizes where any node sits in tensor space. Toggle it from the header to see a live readout as you interact with the graph.
| State | Behavior |
|---|---|
| Idle | Cursor rests at the simplex centroid (balanced COG/EMO/ENV) |
| Hover | Cursor animates toward the hovered node's tensor weights |
| Select | Cursor locks to the selected node's position in the simplex |
| Release | Cursor returns to the selected node (or centroid if none) |
Each node carries numeric COG, EMO, and ENV weights. The cursor position is computed as a barycentric coordinate — cog × top + emo × bottom-left + env × bottom-right — mapping tensor space directly onto the 2D triangle. The dominant vertex pulses, and the cursor color matches the strongest dimension.
Inspiration
Built on the arscontexta methodology — interactive knowledge graphs from conversation context. The OPTX adaptation treats documentation pages as knowledge nodes, cross-references as semantic edges, and AGT tensors as the cognitive navigation layer.