@taskmagic/apps-kavout
v0.0.1
Published
Quantitative equity signals from [Kavout](https://www.kavout.com) — docs: https://www.kavout.com/academy/tutorial
Readme
Kavout
Quantitative equity signals from Kavout — docs: https://www.kavout.com/academy/tutorial
Connection: an API key from the User Center. A key is only issued once your account subscribes to at least one signal in Signal DB — that's the likeliest reason a valid-looking account can't find one.
Kavout takes the key as an api_key query parameter, not a header, so it travels in the
request URL. Avoid pushing raw request URLs into downstream logging.
Base URL: https://www.kavout.com/api — a single endpoint.
One endpoint, two actions
Kavout exposes one URL and selects everything — which signal, which symbols, which date, which columns — by query parameter. So this is two thin actions over one request helper rather than a path per resource:
- Get Signal Data — latest values for the symbols you name, or the whole universe if you name none (a large response).
- Get Signal For Date — the same for a specific
YYYY-MM-DDtrading day, symbols required.
Product ID is the Signal DB product, e.g. kscore_us; you must be subscribed to it.
Rate limit shapes the design
10 requests per minute. That's low enough that a flow looping one call per ticker will hit it quickly, so both actions take a list of symbols and send them in a single request. Pass an array or a comma-separated string — an upstream step's output wires straight in.
No triggers
One request/response endpoint, no subscription, no event feed. "Score changed" would mean storing and diffing history ourselves.
Scope
This piece reads market data. It doesn't place trades, and Kavout's API doesn't offer them.
