---
title: "Transaction Manifests"
path: "/contents/tech/core-protocols/transaction-manifests"
version: "1.2.0"
author: "Hydrate"
createdAt: "2026-02-03T14:58:28.474Z"
updatedAt: "2026-03-16T18:21:00.552Z"
---

# Transaction Manifests

<Infobox>
| **Type** | Transaction Format |
| **Readability** | Human AND machine readable |
| **Execution** | Atomic (all-or-nothing) |
| **Related** | [Blind Signing Problem](/contents/tech/core-concepts/blind-signing-and-transaction-manifests) |
</Infobox>

## Overview

**Transaction manifests** are Radix's instruction format for describing what a transaction should do. Unlike EVM transactions that encode function calls as opaque hex data ([blind signing](/contents/tech/core-concepts/blind-signing-and-transaction-manifests)), manifests are readable scripts that list every operation explicitly.

A manifest reads like a recipe: withdraw XRD from account → deposit into DEX → receive output tokens → deposit into account. Users (and AI agents) can verify every step before signing.

### Composability

Manifests can chain any number of operations into a single [atomic transaction](/contents/tech/core-concepts/atomic-composability). A single manifest might: withdraw collateral from a lending protocol, swap it on a DEX, provide liquidity, and deposit the LP tokens — all atomically. If any step fails, everything reverts.

### AI Agent Compatibility

Manifests are particularly important for [AI agents](/contents/tech/core-concepts/radix-for-ai-agents). An agent can programmatically construct, parse, and verify manifests — understanding exactly what a transaction will do. This is impossible with EVM's opaque calldata.

## External Links

  - [Transaction Manifest Language Guide](/developers/transactions/01-manifest-language)

  - [Transaction Lifecycle](/developers/transactions/02-transaction-lifecycle)

  - [Official Documentation](https://docs.radixdlt.com/)