skylit

A small Python client for Skylit market data. Skylit covers options flow, volatility, dealer positioning (GEX/vanna levels and point-in-time replay) and dark pool, and is built for AI trading agents and the people who build them.

The client has no dependencies, and every call is read-only: it fetches data and never places orders.

pip install skylit
export SKYLIT_API_KEY=...   # create one at https://app.skylit.ai/developer

from skylit import Skylit

client = Skylit()

client.account()                              # balance, prices, limits (free)
client.gex_levels("SPY")                      # king node, gatekeepers, flip, walls
client.historical("SPY", at="2026-09-29T15:30:00Z")   # the board at a past instant
client.vol("iv", ["SPY", "QQQ"])              # Tempest implied volatility
client.vol("cones", "NVDA")                   # expected-move cones
client.flow_tone("TSLA")                      # bull/bear pressure across the chain
client.get("/v1/vol/screener", limit=20)      # any documented endpoint

Responses are the API's JSON, {"data": ..., "meta": ...}. meta carries the remaining credits and the rate-limit state. Errors raise skylit.SkylitError with .status, .code and .message. Failed calls are not charged.

Agents

For Claude, ChatGPT, Cursor, Gemini CLI, LangChain, the OpenAI Agents SDK and other MCP clients, connect to the hosted MCP server at skylit.MCP_URL (https://mcp.skylit.ai/mcp). Setup guides for each client are at SkylitAI/skylit-mcp.

Pricing and terms

Calls are billed in credits (1 credit = $0.001; most calls cost 1 to 5). See the docs. Use of Skylit data is governed by the API Terms, which cover personal and research use; for commercial use, contact support@skylit.ai. The MIT license covers this client's code, not Skylit data.

Staying current

This client and its repo are checked against the live API every 6 hours; new releases follow API changes.