> For the complete documentation index, see [llms.txt](https://puregamma-ai.gitbook.io/puregamma-ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://puregamma-ai.gitbook.io/puregamma-ai/puregamma-ai.md).

# PureGamma AI

One conversation for market context, portfolio impact and research action.

> **One conversation for market context, portfolio impact and research action.**

PureGamma AI is an AI-native investment research and decision-support console for active crypto, equity and options investors. It combines market intelligence, portfolio NAV context, agent research, options analysis, strategy playbooks, simulated backtests, billing entitlements, an OpenAI-compatible API gateway, and iOS/Android apps in one workflow. This documentation set reflects the **2.2 product state (v0.2.x)** and is available in six languages.

#### The three questions we solve

1. **What changed in the market?**
2. **How does it affect my portfolio and risk?**
3. **What research action, scenario or strategy deserves attention now?**

#### What is live today

* Web research console in English and Chinese: Today dashboard, Agent chat with streaming and citations, Private Secretary voice, options research, Backtest Lab, reports and billing.
* iOS and Android apps with Today / Agent / Research / Portfolio / Account surfaces. Research runs, memory controls and trading-safety views are gated by the server capability contract and show an honest "temporarily unavailable" state where the backend endpoint is not yet implemented.
* Server-computed portfolio NAV with explicit freshness windows and stale states, plus read-only connectors: Plaid Investments, Interactive Brokers (OAuth), Hyperliquid public accounts and EVM public wallets.
* Membership tiers: Free / Silver / Gold / Enterprise, with plan fallback (Silver → Pro, Gold → Max).
* OpenAI-compatible API Gateway with a prepaid USD wallet, `sk-pg-…` keys, `GET /v1/models` and `POST /v1/chat/completions`. Routed models include DeepSeek V4 (Pro / Flash), Kimi K3 and GLM 5.2.
* Trading safety foundation: PAPER and SHADOW strategy runtime, trading mandates, kill switches and an immutable ledger. A LIVE trading control plane exists in the backend with **every gate default OFF**; live orders are not enabled.
* MSTR–BTC research dashboard built on official Strategy data through the server-side provider layer.
* Delivery through Email, Telegram, Slack, iOS push and a self-hosted iMessage relay, with idempotent daily-brief delivery.

#### What is being upgraded

* HTTP endpoints for research runs (`/api/research/runs*`) and the memory service (`/api/memory/*`) — the product surfaces already exist behind capability gates; the contract layer ships next.
* Harness deep research and the user memory service remain feature-flagged OFF until end-to-end verification.
* App Store and Google Play release for the mobile apps.
* LIVE trading rollout stays gated behind a real broker adapter, deployment approvals and a pilot cohort; no live order is possible until every gate passes.
* Redis Streams event pipeline and additional data and portfolio connectors.

Availability is always shown in-product. A configured interface, connector or roadmap item is not presented as live data until it has completed end-to-end verification.

#### Documentation map

* [Product Guide](/puregamma-ai/puregamma-ai/product-guide.md) — daily workflows, research, portfolio context and delivery.
* Business plan & product roadmap — customer, product model, distribution and delivery priorities.
* [Platform & Architecture](/puregamma-ai/puregamma-ai/platform-and-architecture.md) — services, data flow, model routing and the trading safety boundary.
* [Trust & Operations](/puregamma-ai/puregamma-ai/trust-and-operations.md) — billing, tiers, gateway wallet, security and operational boundaries.
* [Developer Reference](/puregamma-ai/puregamma-ai/developer-reference.md) — integration and deployment contracts.
* API Gateway — the OpenAI-compatible prepaid API (Chinese) and English version.

```mermaid
flowchart LR
  A["Market, news, options and portfolio sources"] --> B["Research Snapshot"]
  B --> C["Event and asset impact"]
  C --> D["Portfolio impact and research action"]
  D --> E["Web, iOS and Android"]
  D --> F["Reports, alerts and delivery"]
  D --> G["Backtest, PAPER and SHADOW"]
  H["API Gateway (prepaid)"] --> I["DeepSeek, Kimi, GLM"]
  J["LIVE control plane (gates default OFF)"] -.-> K["No live orders enabled"]
```

Start with Getting started, then use the Product Guide for the daily workflow.