DataQuant Connector for Openclaw

1. Install Python 3.8 or newer and the required HTTP dependency:

python -m pip install requests

2. Register for a DataQuant API key at https://app.dataquant.trade.

3. Store the key in an environment variable instead of hard-coding it:

export DATAQUANT_API_KEY="your_dataquant_api_key"

4. Confirm the connection and available allowance:

python scripts/dataquant.py quota --api-key "$DATAQUANT_API_KEY"

5. Run representative DataQuant queries:

python scripts/dataquant.py kline ashare sh600519 --start 2020-01-01 --end 2025-12-31 --adj qfq --api-key "$DATAQUANT_API_KEY"
python scripts/dataquant.py batch ashare sh600519,sz000858 --start 2025-01-01 --adj bfq --api-key "$DATAQUANT_API_KEY"
python scripts/dataquant.py detail ashare sh600519,sz000858 --api-key "$DATAQUANT_API_KEY"
python scripts/dataquant.py screen ashare --min-pe-ratio 0 --max-pe-ratio 30 --min-total-market-cap 1000 --sort chg_20d --api-key "$DATAQUANT_API_KEY"
python scripts/dataquant.py search ashare 600519 --api-key "$DATAQUANT_API_KEY"
python scripts/dataquant.py macro gdp --start 2020 --end 2025 --api-key "$DATAQUANT_API_KEY"

6. Use references/api-reference.md and the online DataQuant API documentation for the complete field definitions, screen allowlist, response examples, package limits, and error-code reference. Batch requests should generally be separated by at least 0.5 seconds, and clients should respect the global 200 requests-per-minute server limit.

Repository files

Market taxonomy

Response and query organization

• K-line responses are organized by symbol and date, and always include adj_factor; supported fields include open, high, low, close, volume, and amount.
• K-line field aliases include o,h,l,c,v,a and their full names. Single-symbol requests default to all fields, while batch requests default to close,volume.
• K-line pagination uses limit and offset; single and batch requests default to 100 rows, subject to the plan's max_single_rows limit.
• Detail responses always retain symbol and date, do not include adj_factor, and default to all fields.
• Screen responses use a fixed 23-column service-defined schema, including symbol, name, market name, date, close, valuation, size, momentum, and moving-average metrics. Filters use min_<column> and max_<column> only for allowlisted columns.
• Search responses contain market, total, count, offset, and a symbols array of code strings; search matches symbol substrings, not Chinese instrument names.
• Macro responses contain a data[] array. The data field is already parsed as an object, so callers do not need a second json.loads operation.
• Quota responses expose remaining allowance for planning request batches.

Price adjustment taxonomy

• bfq: Unadjusted prices and the default mode.
• qfq: Forward-adjusted prices, recommended for indicator calculations to reduce dividend-related gaps.
• hfq: Backward-adjusted prices.
• adj_factor = hfq_close / bfq_close; adjusted prices affect open, high, low, and close, while volume and amount remain unscaled.