# Cortex — Base-Native Agentic Commerce Protocol

> Cortex is a Base-native protocol for agentic commerce. Agents have onchain identity, operate through policy-aware smart accounts, discover merchants and services, accept verifiable quote commitments, record receipts and disputes, and query machine-readable state through REST, MCP, and dashboard interfaces.

## Docs

- [Getting Started](/docs/getting-started.md): Introduction to Cortex and quick start guide
- [Architecture](/docs/architecture.md): System diagram, commerce flow, payment rails, indexed data, and deployment shape
- [Contracts](/docs/contracts.md): Smart contract reference for identity, policy, intents, participant registries, attestations, and commerce
- [REST API](/docs/api.md): Endpoint reference for agents, intents, policies, commerce, analytics, and transaction explanations
- [Local Development](/docs/local-dev.md): Step-by-step guide to running the full stack locally
- [Testnet Deployment](/docs/testnet-deploy.md): Current hosted Base Sepolia/AWS deployment and deployment guide
- [Examples](/docs/examples.md): Merchant onboarding, agent smart accounts, payment policies, quote commitments, receipts, and SDK usage
- [SDK Examples](/docs/sdk-examples.md): TypeScript flows for merchant discovery, quote commitments, payments, x402 policy, receipts, disputes, analytics, and reputation
- [Product Onboarding](/docs/product-onboarding.md): Guided onboarding workspace for merchant profiles, service catalogs, agent policy, and quote acceptance
- [Value Proposition and Use Cases](/docs/value-prop-use-cases.md): Strategic positioning, customer segments, use cases, wedges, product gaps, and messaging
- [Agentic Commerce](/docs/agentic-commerce.md): Merchant registry, service discovery, quote, receipt, dispute, trust-signal, payment-rail, and x402 policy model
- [MCP Server](/docs/mcp.md): Model Context Protocol tools for AI agent integration
- [Design Decisions](/docs/decisions.md): 12 key architectural choices with rationale and tradeoffs
- [Security](/docs/security.md): Threat model covering Base dependencies, intent manipulation, solver abuse, policy bypass, commerce risks, key management, and offchain services

## Optional

- [Full Documentation](/llms-full.txt): Complete documentation in a single file