...
Structured memory cards currently available in the file-backed layer.
ZTRADER MEMORY OS
Decode the Market Matrix.
Command the Z-AI Oracle.
This is not a decorative chat page. It is the central memory layer for research, strategy, signals, execution logic, and module routing across the Ztrader system.
Mode: Central Memory LayerState: File-backed Memory V4.1Engine: Cards / Sources / Logs / Modules
Central research brain, strategy bridge, and memory router.
Memory Library Preview
A central memory layer should show its inventory, not just pretend it knows things.
...
Frameworks, notes, playbooks, and primers connected to memory objects.
...
Latest logged queries written through the query route and dynamic API layer.
Representative Cards
Current top objects available in the local memory spine.
Recent Query Logs
The system should remember what has been asked, not just what looks pretty.
Connected Modules
These modules should eventually read from and write to the same memory spine instead of behaving like separate little kingdoms.
Research
Macro Research
Cross-asset macro notes, event frameworks, rates, FX, commodities, and geopolitical stress mapping.
Derivatives
Options Lab
Volatility structure, options playbooks, earnings setups, skew, term structure, and scenario mapping.
Execution
Order Flow
Execution logic, liquidity windows, bid/ask microstructure, flow interpretation, and forced unwind signals.
Event
Earnings
Earnings framework, gap logic, post-print behavior, positioning, and cross-sector sensitivity analysis.
Charts
Technical Charts
Structure phase, momentum bias, candle context, support/resistance, and annotated market map generation.
Quant
Quant Engine
Factor logic, signal definitions, backtest modules, ranking engines, and screening workflows.
Strategy
Strategy Pod
CTA, macro, event, execution, and hybrid strategy pods linked to regime-aware memory objects.
Risk
Risk Monitor
Shock regime handling, position discipline, execution risk, scenario stress, and portfolio constraint logic.
Recent Memory Objects
Cards are the useful unit. Not just random chunks floating in a vector swamp.
Oil Shock Framework
Hormuz chokepoint, logistics shock, physical supply repricing, and macro spillover path.
Dual-Blade CTA
Direction filter first, vulnerable liquidity window second. Trade only when both blades align.
USDJPY Basis Tension
Dollar funding stress, basis pressure, intervention risk, and cross-border balance sheet tension.
Earnings Gap Playbook
Pre-print setup, implied move, post-gap behavior, and failure patterns after the first hour.
System Architecture Preview
The future RAG core should not be “chat only”. It should route between retrieval, graph, structured data, and tool modules.
User Query
↓
Intent Router
↓
Retrieval + Graph + Data
↓
Context Builder
↓
Answer / Module Launch
Memory Layer
Documents, chunks, cards, graph edges, citations, and structured research objects.
Query Router
Ask, research, strategy, data, and module-launch requests are routed through different retrieval paths.
Module Bridge
Macro, options, order flow, quant, charts, and risk tools should all consume one shared memory spine.