Reference

CLI reference

Commands and flags for connecting, running backtests, starting live deployments, and research workflows.

AlphaLens Core docs

The alphalens command is installed by alphalens-core.

bash

pip install "alphalens-core[live,cloud,fast]"

`alphalens connect`

Cache an AlphaLens API key locally:

bash

alphalens connect --api-key alens_...

You can also omit the flag and paste the key when prompted:

bash

alphalens connect

After connecting, future runs can use the cached session. You can also skip this step and set ALPHALENS_API_KEY for each process.

Run one backtest:

bash

alphalens run --strategy my_strategy:MyStrategy

Flag	Purpose
`--strategy`	Required Python import path in `module:ClassName` format
`--start`	Override `class.start`
`--end`	Override `class.end`
`--universe`	Override symbols with `SPY,QQQ,IWM` or `@symbols.txt`
`--resolution`	Override `class.resolution`, such as `1day` or `5minute`
`--param key=value`	Pass strategy parameters, repeatable
`--cash`	Override `class.initial_cash`
`--benchmark`	Override `class.benchmark_symbol`
`--report`	Write local HTML report
`--save-equity`	Write equity curve CSV
`--save-trades`	Write trades CSV
`--save-orders`	Write orders CSV
`--log-to`	Write trade and equity log to SQLite
`--plot`	Show a local matplotlib equity chart
`--no-track`	Skip local run registry tracking
`--no-progress`	Suppress monthly progress output
`--runs-db`	Override local run registry DB

Run a strategy against Alpaca with completed bars:

bash

alphalens live --strategy my_strategy:MyStrategy

Flag	Purpose
`--strategy`	Required Python import path in `module:ClassName` format
`--broker`	Broker backend, currently `alpaca`
`--data`	`alphalens` hosted data or direct `polygon`
`--live`	Route to live Alpaca instead of paper
`--history-bars`	Bars seeded before live processing
`--history-max-bars`	Maximum rolling bars retained
`--lookback-days`	Tail lookback used by market data polling
`--poll-sec`	Poll interval while waiting for completed bars
`--idle-log-sec`	Idle waiting log cadence
`--state-path`	Checkpoint path
`--allow-open-orders`	Start even if Alpaca reports open orders
`--continue-on-error`	Keep polling after callback errors
`--deployment-id`	Reuse an existing Strategy Center deployment
`--deployment-name`	Display name for the deployment
`--no-telemetry`	Disable AlphaLens telemetry writes
`--param key=value`	Pass strategy parameters, repeatable

Use direct Polygon data only when you manage your own POLYGON_API_KEY:

bash

alphalens live --strategy my_strategy:MyStrategy --data polygon

Run walk-forward analysis:

bash

alphalens wfa \
  --strategy my_strategy:MyStrategy \
  --train 504 \
  --test 63 \
  --step 63

Walk-forward and batch workflows are research tools. Strategy Center v1 focuses on synced backtests and live deployments.