Freqtrade strategy robustness check is a search pattern for a reason: traders want a checklist they can share, not a vague essay. This page lists seven items in order. It maps to freqtrade strategy ready to deploy intent and complements When is a strategy ready to deploy?.

1. Data quality and bar integrity

Bad bars create fake alpha. Confirm gaps, timezone alignment, and feed consistency (DQG).

2. Costs: fees, spread, slippage

Freqtrade backtest slippage fees must match what you can execute. Stress-test costs (Cost drag).

3. No lookahead in signals

Indicators must be computed bar-by-bar without future leakage (Look-ahead bias).

4. Hyperopt honesty

If you used hyperopt, winners are candidates until forward validation (Hyperopt).

5. Walk-forward or sequential OOS

At least one honest time-split protocol (WFA setup, What is WFA?).

6. Parameter sensitivity

Not a single needle in parameter space (Parameter sensitivity).

7. Second opinion via Kiploks

Export artifacts for structured verdicts and robustness metrics (Integration).

Add-on: logging you will actually read

Before live, verify you can answer these questions from logs alone:

why each trade opened and closed
which config version was active
which data file or download revision was used

If logs are noisy or incomplete, you will debug live incidents by astrology.

Add-on: define "done" for each checklist item

Turn each checkbox into a pass/fail definition:

DQG: list the checks you ran and the anomalies you accepted or rejected
costs: show three cost tiers and which strategies survive each
WFA: show window count and parameter stability summary

A checklist without definitions becomes theater.

If you fail any item

Stop sizing up. Fix the weakest link first. Freqtrade drawdown too high in live is often a symptom of skipped steps 1-4.