FragmentAPI/docs/getting-started/configuration.md

Library and Configuration

Main entry point of the library is FragmentClient.

FragmentClient(
    seed: str,
    api_key: str,
    cookies: dict[str, Any] | str,
    wallet_version: str = "V5R1",
    timeout: float = 30.0,
)

Parameters

• seed: wallet mnemonic (12, 18, or 24 words)
• api_key: Tonapi key from https://tonconsole.com
• cookies: Fragment cookies as a dictionary or JSON string
• wallet_version: "V4R2" or "V5R1"
• timeout: request timeout in seconds

If api_key is too short or cookies are incomplete, initialization fails immediately.

Required cookies

• stel_ssid
• stel_dt
• stel_token
• stel_ton_token

Minimal initialization pattern

from pyfragment import FragmentClient

client = FragmentClient(
    seed="word1 word2 ... word24",
    api_key="YOUR_TONAPI_KEY",
    cookies={
        "stel_ssid": "...",
        "stel_dt": "...",
        "stel_token": "...",
        "stel_ton_token": "...",
    },
)

Use it inside async context manager:

async with client:
    wallet = await client.get_wallet()

You can also create the client directly inside async with if you prefer one-block setup.

Validation behavior

At initialization, library validates:

• seed format,
• cookie shape and required keys,
• supported wallet version,
• parseability of cookie JSON strings.

Constructor-level issues are raised as ConfigurationError or CookieError.

Tip: keep validation failures visible in logs during initial integration. They save a lot of debugging time.