> ## Documentation Index
> Fetch the complete documentation index at: https://docs.timepointai.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Introduction
> Timepoint AI — infrastructure for temporal reasoning. Render the past. Simulate the future. Score the predictions. Accumulate the graph.
# Timepoint AI
Timepoint is the infrastructure layer for temporal reasoning. We render the past to understand it. We render the future to choose it.
## What is Timepoint?
Timepoint is an open-source ecosystem of integrated services that work together to build a persistent, growing model of historical and future causality:
**The Operator** — the natural-language interface to the whole stack. Ask it in plain language to find money, run a simulation, render a moment, or explore the graph; it calls the services and reports credits as it goes. The authenticated home of app.timepointai.com, available over REST and MCP.
**The Front Door** — API gateway at api.timepointai.com. Handles authentication (Apple, Google, GitHub OAuth), JWT signing, credits, and user management. All client traffic enters here.
**The Reality Writer** — type any moment in history, get a grounded scene with characters, dialog, and a photorealistic image via a 14-agent critique-loop pipeline.
**The Temporal Causal Graph** — tens of thousands of historical moments connected by hundreds of thousands of typed causal edges, spanning 700 BCE to 2026 and growing 24/7. PostgreSQL-backed directed graph that accumulates rendered past and simulated future into a growing Bayesian prior. (Live counts: `GET /api/v1/stats`.)
**The Commerce Layer** — Stripe payments, Apple IAP, and invite codes. Coordinates with Gateway to grant credits.
**The Data Format** — JSON-LD interchange format with content-addressed SHA-256 hashing. The universal data contract.
**The Quality Certifier** — temporal reasoning benchmark with 60 adversarial tasks, 5 scoring axes, designed to stay hard through 2030.
**The Settlement Layer** — prediction market protocol on BASE L2 where rendered futures are scored against reality.
## Core Concept: SNAG
**Social Network Augmented Generation** synthesizes and maintains structured social graphs — with causal provenance, knowledge flow, emotional states, and temporal consistency — to ground LLM generation in complex group dynamics.
Where **RAG** answers questions from what was written down, **SNAG** reasons about what people *did*, *felt*, and *caused*.
| Aspect | RAG | SNAG |
| ------------- | ------------------ | ----------------------------------------- |
| **Input** | Document corpus | Scenario + entities + relationships |
| **Structure** | Flat chunks | Directed causal graph |
| **Temporal** | Snapshot | Multi-mode (forward, branching, cyclical) |
| **Output** | Answer + citations | Simulation state + dialog + provenance |
## Architecture
All client traffic enters through the API Gateway (api.timepointai.com), which handles authentication, credits, and user management. The Gateway proxies generation requests to Flash with an X-User-ID header. Flash produces grounded scenes, drawing on Clockchain (causal graph) and SNAG-Bench (benchmark). Billing handles payments and coordinates with Gateway to grant credits. Flash and Clockchain share TDF (data format) as their interchange layer.
```
api.timepointai.com
│
┌─────────▼──────────┐
│ Gateway (Auth, │
│ Credits, Users) │
└─────────┬──────────┘
│ X-User-ID
┌─────────▼──────────┐ ┌─────────────────┐
│ Flash (Renderer) │ │ Billing │
│ (no auth — pure │ │ (Stripe, IAP, │
│ generation engine)│ │ invite codes) │
└─────────▲──────────┘ └────────┬────────┘
│ │
┌─────────┴───────────┐ │
│ │ credits/grant
┌────────┴────────┐ ┌─────────┴───────┐
│ Clockchain │ │ SNAG-Bench │
│ (Causal Graph) │ │ (Benchmark) │
└────────┬────────┘ └─────────────────┘
│
┌───────▼───────┐
│ TDF │
│ (Data Format) │
└───────────────┘
```
## Get Started
New to Timepoint? These four pages cover the shortest path from zero to a working integration:
Clone, install, and render your first moment — one click from here to a working `curl` command.
The four credential types (Bearer JWT, API Key, X-Service-Key, X-Admin-Key) and when to use each.
The MCP tools (`clockchain_stats`, `clockchain_search`, `clockchain_moment`, `clockchain_neighbors`, `flash_render`).
HTTP status codes, rate-limit headers, and a decision tree for common failures.
### Go Deeper
Timepoints, temporal modes, and the Clockchain
Gateway, Flash, and Clockchain API endpoints
All source code — Apache-2.0 licensed
WCAG 2.1 Level A conformance statement