# Regent Protocol Documentation

> Solana-native trust and auditability infrastructure for AI agents in regulated finance. Regent gives an AI agent a verifiable identity (AgentID) bound to a KYC-verified responsible party, bounds its authority with on-chain spending mandates, and anchors every action to Solana as a tamper-evident audit trail.

Every page below links to its markdown version (append `.md` to any docs URL, or send `Accept: text/markdown`). The whole corpus in one file: [llms-full.txt](https://docs.regentprotocol.org/llms-full.txt). Agent skill definition: [skill.md](https://docs.regentprotocol.org/skill.md).

## Overview

- [Introduction](https://docs.regentprotocol.org/index.md): Solana-native trust and auditability infrastructure for AI agents in regulated finance.

## Core Concepts

- [AgentID](https://docs.regentprotocol.org/concepts/agent-id.md): Machine-verifiable identity for AI agents, KYC-bound to a responsible party and anchored on Solana.
- [Mandate](https://docs.regentprotocol.org/concepts/mandate.md): On-chain spending rules that bound what an agent can do without requiring per-action human approval.
- [Audit Trail](https://docs.regentprotocol.org/concepts/audit-trail.md): Every meaningful agent action hashed, Merkle-batched, and anchored on Solana. Tamper-evident by construction.
- [Guardian](https://docs.regentprotocol.org/concepts/guardian.md): Real-time anomaly detection on agent behavior, with explainable risk scores and threshold alerts.
- [Responsible Party](https://docs.regentprotocol.org/concepts/responsible-party.md): The human or entity legally accountable for an agent's actions. Every AgentID is bound to one.

## Quick Start

- [Three Steps](https://docs.regentprotocol.org/quickstart/three-steps.md): From zero to a first authorized agent action in three steps.

## Regent Control

- [Overview](https://docs.regentprotocol.org/control.md): Runtime policy enforcement for AI agent actions — allow / deny / escalate, in the data path.
- [Onboarding Guide](https://docs.regentprotocol.org/control/onboarding.md): Onboard an org onto Regent Control end-to-end via the dashboard, then deploy the sidecar.
- [Run the Sidecar](https://docs.regentprotocol.org/control/run-sidecar.md): Bring the Regent Control sidecar up with Docker — a zero-prod demo and a real deployment.
- [Walkthrough: Refund Agent](https://docs.regentprotocol.org/control/refund-agent.md): A bank customer-support refund agent, gated end-to-end by Regent Control — role, spend caps, delegation, and human approval.
- [SDK & Integration](https://docs.regentprotocol.org/control/integration.md): Wire an agent's actions through Regent Control — Python / TypeScript SDKs, MCP, the error contract, and the token verifier.

## Cloud Gateway & MCP

- [Overview](https://docs.regentprotocol.org/gateway.md): Point an AI agent at one URL and one key — every tool call gated, credentials in Regent's vault. Set up by chatting over MCP.
- [Admin MCP (setup)](https://docs.regentprotocol.org/gateway/admin-mcp.md): Onboard agents, mandates, keys and tools by chatting — every Admin MCP tool with a use case.
- [Tools MCP (runtime)](https://docs.regentprotocol.org/gateway/tools-mcp.md): How an agent calls governed tools + pays from its wallet over the Cloud Gateway's MCP endpoint.
- [Tool Catalog](https://docs.regentprotocol.org/gateway/tool-catalog.md): Every tool an agent can use through the Cloud Gateway — with a governance use case for each — plus how to add your own.

## Python SDK

- [Introduction](https://docs.regentprotocol.org/sdk-python/guide/introduction.md): Async Python SDK for the Regent Protocol — agents, mandates, audit, guardian.
- [Installation](https://docs.regentprotocol.org/sdk-python/guide/installation.md): Install the Regent Python SDK and verify the version.
- [Authentication](https://docs.regentprotocol.org/sdk-python/guide/authentication.md): API keys, the gateway base URL, and how organization scoping works.

## REST API

- [Authentication](https://docs.regentprotocol.org/rest-api/authentication.md): Authenticate REST API requests with a bearer token.
- [Agents](https://docs.regentprotocol.org/rest-api/agents.md): Register, list, get, and revoke agents over REST.
- [Mandates & Authorize](https://docs.regentprotocol.org/rest-api/mandates.md): Create spending mandates and authorize payments against them — the pre-execution gate.
- [Audit Events](https://docs.regentprotocol.org/rest-api/audit.md): Ingest tamper-evident audit events and verify their Merkle proofs.
- [Error Reference](https://docs.regentprotocol.org/rest-api/errors.md): Every error code the API returns, what it means, and what to do.

## Solana Programs

- [Overview](https://docs.regentprotocol.org/solana-programs/overview.md): The three Anchor programs that anchor Regent Protocol on Solana.
- [AgentRegistry](https://docs.regentprotocol.org/solana-programs/agent-registry.md): On-chain agent identity registration and revocation.
- [MandateRegistry](https://docs.regentprotocol.org/solana-programs/mandate-registry.md): On-chain mandate registration and revocation.
- [AuditAnchor](https://docs.regentprotocol.org/solana-programs/audit-anchor.md): Merkle batch root storage for audit events.
- [Verifying on-chain](https://docs.regentprotocol.org/solana-programs/verification.md): End-to-end recipes for independently verifying Regent agents, mandates, and audit events against Solana.

## Use Cases

- [AI Trading Agent](https://docs.regentprotocol.org/use-cases/ai-trading-agent.md): A real AI agent trading BTC on Binance under Regent Protocol governance — identity, mandate, audit, kill switch.

## Resources

- [GitHub repositories](https://docs.regentprotocol.org/resources/github.md): Where the code lives.
- [Live demo](https://docs.regentprotocol.org/resources/demo.md): Try Regent Protocol end-to-end without writing any code.
- [Agent Validation](https://docs.regentprotocol.org/resources/agent-validation.md): agentvalidate — the joint Regent × RagMetrics trust layer: who an agent is, who answers for it, and how well it behaves.