CLE CLI

Getting Started

Creating Your First CLE Project

With NPM:

npm create cle@latest

With Yarn:

yarn create cle@latest

With PNPM:

pnpm create cle@latest

Then follow the prompts!

To create a uniswapprice template, you can directly specify the project name and the desired template using additional command line options. For instance, you can run the following command:

# npm 6.x
npm create cle@latest my-cle-uniswapprice --template uniswapprice

# npm 7+, extra double-dash is needed:
npm create cle@latest my-cle-uniswapprice -- --template uniswapprice

# yarn
yarn create cle@latest my-cle-uniswapprice --template uniswapprice

# pnpm
pnpm create cle@latest my-cle-uniswapprice --template uniswapprice

CLI

Note: unsafe (define unsafe: true in the cle.yaml) means the CLE is compiled locally and only contains partial computation (so that proving and executing will be faster).

The workflow of local CLE development must follow: Develop (code in /src) -> Compile (get compiled wasm image) -> Execute (get expected output) -> Prove (generate input and pre-test for actual proving in zkOracle) -> Verify (verify proof on-chain).

To upload and publish your CLE, you should Upload (upload code to IPFS), and then Publish (register CLE on onchain CLE Registry).

Compile

Compile for Full Image (Link Compiled with Compiler Server).

Usage

cle compile [root]

Options

Options	Description
--yaml-path <path>	Path to yaml file
--dir <path>	Path to directory containing cle.yaml

Execute

Execute Full Image.

Please save the CLE_STATE_OUTPUT string for following prove steps.

Usage

cle exec [...params] [root]

Usage cases

cle exec <blockId> [root]
cle exec <blockId> <offchainData> [root]

Arguments

Arguments	Description
<block id>	Block number (or block hash) as runtime context
<offchainData>	offchain data

Set Up

Set Up Full Image

• circuit-size: Specify the circuit size of image instead of the default recommended. eg. cle setup -- --circuit-size <size> (eg. 22).

Usage

cle setup [root]

Options

Options	Description
-k, --circuit-size <size>	Circuit size (k in 2^k) of image

Prove

Prove Full Image

Usage

cle prove [...params] [root]

Usage cases

cle prove <blockId> <expectedStateStr> [root]
cle prove <blockId> <offchainData> <expectedStateStr> [root]

Arguments

Arguments	Description
<block id>	Block number (or block hash) as runtime context
<expected state>	State output of the CLE execution
<offchainData>	offchain data

Options

Options	Description
-i, --inputgen	Run in input generation Mode
-t, --test	Run in test Mode
-p, --prove	Run in prove Mode

Upload

Upload CLE (Code and Full Image).

Please save the ipfs_hash from the output dialog for following publish steps.

Usage

cle upload [root]

Verify

Verify Proof Onchain.

Usage

cle verify <prove task id>

Arguments

Arguments	Description
<prove task id>	Task id of prove task

Publish

Publish and Register CLE Onchain.

See also: Verifier Contract Interface.

Usage

cle publish <ipfs_hash> [bounty_reward_per_trigger]

Arguments

Arguments	Description
<ipfs hash>	IPFS hash of uploaded CLE
[bounty reward per trigger]	Bounty reward per trigger in ETH

Deposit

Publish and register CLE Onchain.

Usage

cle deposit <deployed contract address> <deposit amount>

Arguments

Arguments	Description
<deployed contract address>	Contract address of deployed verification contract address
<deposit amount>	Deposit amount in ETH

Config

Configuring CLE

When running CLE from the command line, the tool will automatically attempt to locate a configuration file named cle.config.js in the project’s root directory. It also supports other file extensions such as JS and TS.

The most basic config file looks like this:

// cle.config.js
export default {
  // config options
}

You can also explicitly specify a config file to use with the --config CLI option (resolved relative to cwd):

cle --config my-config.js

Config Intellisense

Since CLE ships with TypeScript typings, you can leverage your IDE's intellisense with jsdoc type hints:

/** @type {import('@ora-io/cle-cli').UserConfig} */
export default {
  // ...
}

Alternatively, you can use the defineConfig helper which should provide intellisense without the need for jsdoc annotations:

import { defineConfig } from '@ora-io/cle-cli'

export default defineConfig({
  // ...
})

CLE also directly supports TS config files. You can use cle.config.ts with the defineConfig helper as well.

Config Options

JsonRpcProviderUrl

• Type: object
• Default: { mainet: "", sepolia: "", goerli: ""}

Update your Etherum JSON RPC provider URL here. It is recommended to use providers that support debug_getRawReceipts RPC method.

UserPrivateKey

• Type: string

Update your private key here to sign messages & send txs. The CLI will inform you before sending out any tx. The cle.config.ts is in .gitignore by default. CLI will never upload / disclose your private key by default.

ProverProviderUrl

• Type: string
• Default: https://rpc.zkwasmhub.com:8090

CompilerServerEndpoint

• Type: string
• Default: https://compiler.ora.io/compile/

PinataEndpoint

• Type: string
• Default: https://api.pinata.cloud/pinning/pinFileToIPFS

PinataJWT

• Type: string

WasmBinPath

• Type: string
• Default: [root]/build/cle.wasm

CLE CLI Build-In a tag name is root.
The root is user project root path.
Of course, you can also place this tag at any position within the string.