SnapTrade Trading Skill#

SnapTrade connects OpenClaw to the user's brokerage (IBKR, Questrade, Alpaca, and more) through a unified API. This skill handles all account data retrieval and order execution. For market price data and OHLCV bars, use the massive-data-feed skill — SnapTrade does not provide chart data.

---

Setup#

Run once before first use:

bash

bash scripts/setup.sh
source .venv-snaptrade/bin/activate

See README.md for full onboarding — how to get credentials, register a user, and connect a brokerage account.

---

SDK Initialization#

Always run this first before any other call:

python

import os
from snaptrade_client import SnapTrade

snaptrade = SnapTrade(
    consumer_key=os.environ["SNAPTRADE_CONSUMER_KEY"],
    client_id=os.environ["SNAPTRADE_CLIENT_ID"],
)
user_id = os.environ["SNAPTRADE_USER_ID"]
user_secret = os.environ["SNAPTRADE_USER_SECRET"]

---

What to do next — read the right reference file#

Each reference file covers one domain. Read only what you need.

TaskReference fileGet accounts, balances, positions, orders, historical valuereferences/account-data.mdResolve a ticker to a symbol ID + get quotesreferences/symbol-resolution.mdPlace an equity trade (limit, market, bracket)references/place-orders.mdPlace an options order (single or multi-leg)references/options-trading.mdPlace a crypto orderreferences/crypto-trading.mdCancel an order or refresh holdingsreferences/cancel-refresh.mdGet historical transactions / activity logreferences/historical-data.md

---

Key constraints to keep in mind#

• trade_id from an impact check expires in 5 minutes — place the order immediately after getting it, don't do other work in between
• SnapTrade does not provide OHLCV or price history — use massive-data-feed for that
• Not all brokerages support trading — some are data-only connections
• Always resolve symbol IDs fresh — don't cache them long term, they can change
• After any order placement or cancellation, trigger a manual refresh so the account state stays current (see references/cancel-refresh.md)