tinyman-py-sdk

Tinyman Python SDK

Design Goal

This SDK is designed for automated interaction with the Tinyman AMM. It will be most useful for developers who wish to create automated trading programs/bots. It may also be useful to create an alternative UI but that is not a design goal of this library. It is designed to be reasonably low level so that pieces can be used in isolation.

Status

This SDK is currently under active early development and should not be considered stable.

The SDK supports Tinyman V2 and V1.1.

Requirements

• Python 3.8+
• py-algorand-sdk 1.10.0+

Installation

tinyman-py-sdk is not released on PYPI. It can be installed directly from this repository with pip:

pip install git+https://github.com/tinymanorg/tinyman-py-sdk.git

Integration

If you are integrating your project into Tinyman, you can provide client_name while setting up Tinyman Client classes. The client name will be added to the application call transaction's note field. It is recommended and completely optional.

client = TinymanV2MainnetClient(..., client_name="project name", ...)

V2

Sneak Preview

# examples/v2/sneak_preview.py

from examples.v2.utils import get_algod
from tinyman.v2.client import TinymanV2TestnetClient

algod = get_algod()
client = TinymanV2TestnetClient(algod_client=algod)

# Fetch our two assets of interest
USDC = client.fetch_asset(10458941)
ALGO = client.fetch_asset(0)

# Fetch the pool we will work with
pool = client.fetch_pool(USDC, ALGO)
print(f"Pool Info: {pool.info()}")

# Get a quote for a swap of 1 ALGO to USDC with 1% slippage tolerance
quote = pool.fetch_fixed_input_swap_quote(amount_in=ALGO(1_000_000), slippage=0.01)
print(quote)
print(f"USDC per ALGO: {quote.price}")
print(f"USDC per ALGO (worst case): {quote.price_with_slippage}")

Tutorial

You can find a tutorial under the examples/v2/tutorial folder.

To run a step use python <file_name> such as python 01_generate_account.py.

Prerequisites

1. Generating an account
2. Creating assets

Steps

3. Bootstrapping a pool
4. Adding initial liquidity to the pool
5. Adding flexible (add two asset with a flexible rate) liquidity to the pool
6. Adding single asset (add only one asset) liquidity to the pool
7. Removing liquidity to the pool
8. Removing single asset(receive single asset) liquidity to the pool
9. Swapping fixed-input
10. Swapping fixed-output

Example Operations

V1.1

Sneak Preview

from tinyman.v1.client import TinymanTestnetClient
from algosdk.v2client.algod import AlgodClient

algod = AlgodClient('<TOKEN>', 'http://localhost:8080', headers={'User-Agent': 'algosdk'})
client = TinymanTestnetClient(algod_client=algod)

# Fetch our two assets of interest
TINYUSDC = client.fetch_asset(21582668)
ALGO = client.fetch_asset(0)

# Fetch the pool we will work with
pool = client.fetch_pool(TINYUSDC, ALGO)

# Get a quote for a swap of 1 ALGO to TINYUSDC with 1% slippage tolerance
quote = pool.fetch_fixed_input_swap_quote(ALGO(1_000_000), slippage=0.01)
print(quote)
print(f'TINYUSDC per ALGO: {quote.price}')
print(f'TINYUSDC per ALGO (worst case): {quote.price_with_slippage}')

# See the examples for the rest...

Conventions

• Methods starting with fetch_ all make network requests to fetch current balances/state.
• Methods of the form prepare_X_transactions all return TransactionGroup objects (see below).
• All asset amounts are returned as AssetAmount objects which contain an Asset and amount (int).
• All asset amount inputs are expected as micro units e.g. 1 Algo = 1_000_000 micro units.

Signing & Submission

The SDk separates transaction preparation from signing and submission to leave the developer in full control of how transactions are signed and submitted to the network.

Preparation

The prepare_X_transactions methods all return a TransactionGroup object. This is a container object containing a list of transaction objects (.transactions) and a list for signed transactions (.signed_transactions).

transaction_group = client.prepare_app_optin_transactions(account['address'])

Signing

In most cases some of the transactions have a corresponding entry in .signed_transactions because they have been signed by the Pool LogicSig. The remaining transactions should be signed by the 'user'.

The TransactionGroup includes a method to do this when signing with a private key:

transaction_group.sign_with_private_key(account['address'], account['private_key'])

This helper method is equivalent to the following:

for i, txn in enumerate(transaction_group.transactions):
    if txn.sender == account['address']:
        transaction_group.signed_transactions[i] = txn.sign(account['private_key'])

Any alternative method of signing can be used here following the same pattern. For example using KMD:

kmd = algosdk.kmd.KMDClient(KMD_TOKEN, KMD_ADDRESS)
handle = kmd.init_wallet_handle(KMD_WALLET_ID, KMD_WALLET_PASSWORD)
for i, txn in enumerate(transaction_group.transactions):
    if txn.sender == account['address']:
        transaction_group.signed_transactions[i] = kmd.sign_transaction(handle, KMD_WALLET_PASSWORD, txn)

A User account LogicSig can also be used in a similar way or using the sign_with_logicsig convenience method:

transaction_group.sign_with_logicsig(logicsig)

Submission

A TransactionGroup containing fully signed transactions can be submitted to the network in either of two ways:

Using an Algod client:

algod = AlgodClient(TOKEN, ADDRESS, headers={'User-Agent': 'algosdk'})
txid = algod.send_transactions(transaction_group.signed_transactions)

Or, using the convenience method of the TinymanClient:

result = client.submit(transaction_group, wait=True)

This method submits the signed transactions and optionally waits for confirmation.

License

tinyman-py-sdk is licensed under a MIT license except for the exceptions listed below. See the LICENSE file for details.

Exceptions

tinyman/v1/asc.json is currently unlicensed. It may be used by this SDK but may not be used in any other way or be distributed separately without the express permission of Tinyman.