meow-meow-hood 0.2.2

A robinhood api wrapper for option data

Meow-Meow-Hood API

A Robinhood API wrapper for fast option market data

Installation

uv add meow-meow-hood
pip install meow-meow-hood

Why?

If you're using Robinhood as a broker still...
And if you want to view option market data at a reasonable speed. This was made for fun.

Authentication

Ensure you are logged in locally to Robinhood on either:

• Chrome
• Firefox

It works by extracting a locally stored access token from your browser data.
By default the client automatically checks Chrome first and then Firefox for a valid token.
If a stored token is rejected and open_browser=True, the refresh helper may briefly open the installed browsers to refresh auth state.
This behavior can be disabled on class creation with extract_token=False.
You will need to pass in the access token manually if token extraction is disabled.
Config folder is created at the cwd labeled .meow-meow-hood.
If cache enabled db file is placed inside this folder. Same for extract_token.

Robinhood(extract_token=False, access_token="...")

Usage

JSON responses are returned as named data classes for easier parsing.
Examples: FullQuote, OptionInstrument, OptionGreekData, etc.\

@dataclass(frozen=True, slots=True)
class FullQuote(ApiPayloadMixin):
    ask_price: float
    ask_size: int
    bid_price: float
    bid_size: int
    ...

(Refer to robinhood/api_dataclasses.py for full implementation details.)

OptionRequest is the main class when requesting option data.

OptionRequest(
    *,
    symbol: str,
    exp_date: str | None = None,
    option_type: Literal['call', 'put'] | None = None,
    strike_price: float | None = None
)

Example:

# With context manager
with Robinhood() as rh:
  spy_dates: list[str] = rh.get_expiration_dates("SPY")
  spy_request1 = OptionRequest(symbol="SPY", exp_date=spy_dates[0])
  spy_option_data = rh.get_option_greeks_batch_request(spy_request1)
  for option_request, options in spy_option_data.items():
    print(option_request, len(options))

# No context manager
rh = Robinhood()
spy_quote: FullQuote = rh.get_stock_quotes("SPY")
print(spy_quote.ask_price)
rh.close()

Local caching

This library uses a local SQLite database to cache option instruments and reduce the amount of requests made per call.
Cache is validated with a TTL of the next day at 9:30 EDT.
See /docs/design_notes.md for more details

Example where caching improves speed.

with Robinhood(enable_cache = True) as rh:
  dates = rh.get_expiration_dates("SPY")
  strike_map = rh.get_strike_prices(symbol="SPY",exp_date=dates[0])
  call_request = OptionRequest(
    symbol="SPY",
    exp_date=dates[0],
    option_type="call",
  )
  strikes = strike_map[call_request]
  spy_option_list = []
  # Generate a list of OptionRequests you want to view
  for s in strikes:
    opt_req = OptionRequest(symbol="SPY",exp_date=dates[0],strike_price=s)
    spy_option_list.append(opt_req)

  while True:
    option_data = rh.get_option_greeks_batch_request(spy_option_list)
    for option_request, options in option_data.items():
      option = options[0]
      print(option_request.strike_price, option.ask_price, option.bid_price)
    # Delay should be added to avoid rate limit
    time.sleep(0.25)

Trust Me Bro Benchmarks

Benchmark summary for 10 symbols¹ and the second nearest expiration date across 10 runs(~1.6k options returned per run).

Metric	Cold Cache	Warm Cache	Improvement
Average get_expiration_dates time	0.06957s	0.00013s	99.81%
Average get_strike_prices time	0.31505s	0.00029s	99.91%
Average get_option_greeks_batch_request time	3.48793s	1.18787s	65.94%
Average total time per run	7.33413s	1.19204s	83.75%

Cold cache represents a run without any cached option data. Warm cache represents a run with cache hits.

TODO log

See /docs/todo.md for planned features.

1. SPY, TSLA, QQQ, NVDA, GOOG, MSFT, AMZN, TSM, META, JPM ↩