Paper mode first: test Breakflare trading agent before live keys

Run Breakflare Trading agent as a simulator on your shortlist until the fill log matches your written gates. Keep live Binance keys off until that checklist is boring. Breakflare is a daily hype-momentum tracker for CoinGecko ranks 51-200. It scores hype, 3d/7d momentum, Climb, Breakout Score™, and confidence. Core research does not place trades or publish price targets. Trading agent starts in paper by default.

Key Takeaways

• Paper mode lets you rehearse score floors, size caps, stops, and a kill switch without exchange risk or credit spend.
• Keep paper on until you can explain every fill and every skip from the rule text.
• Do not paste live trade-only keys just to see if anything happens overnight.
• Live later means keys + PAYG credits at 0.5% of fill notional; research scores still are not price targets.

What is paper mode allowed to do?

Paper mode is allowed to:

• Read the same daily research fields you see on the terminal
• Simulate entries and exits under your configured rules
• Log skips, orders, and notional paper fills for review
• Run while you learn the scoring stack

Paper mode is not allowed to:

• Place real exchange orders
• Debit PAYG credits
• Prove that tomorrow’s mid-caps will follow last week’s paper path
• Substitute for the human UTC shortlist habit

Think of paper as a flight simulator attached to a research instrument. The instrument still measures relative attention and breakout layers. The simulator asks whether your automation personality matches that instrument. For product context, see the trading agent introduction.

The highest-value paper days are often the quiet ones: lots of skips, few fills, gates doing their job. Chatty paper logs usually mean your floors are theater.

How do you map Breakout Score™ gates onto paper entries?

Write gates in the same language as the board:

1. Universe: ranks 51-200 only (respect exclusions such as stables).
2. Primary sort: Breakout Score™ floor for entry eligibility.
3. Confidence floor: refuse exploratory thin-coverage rows (confidence guide).
4. Momentum check: require accelerating or at least non-rolling 7d hype momentum when that is part of your thesis.
5. Climb (optional): add rank climb only if you truly use it on the desk.
6. Size cap: max notional per name and per day.
7. Stops / max hold: define exit mechanics before entry excitement (max-hold and stop rules).
8. Kill switch: daily loss or error budget that halts new live-intent behavior (kill switch drill).

Then run paper for enough sessions that luck stops dominating the story. One lucky afternoon is not a graduation ceremony. Cross-check paper candidates against false-breakout patterns you already know (hype fades before rank).

If alerts are on, bind them to the same floors (Telegram setup). Human pings and agent rules should not disagree by design.

How do you read the activity log without P&L theater?

Open the log looking for rule fidelity, not for a victory chart.

Useful questions:

• Why was this name eligible under Breakout Score™ and confidence?
• Why did the agent skip a loud social ticker you secretly liked?
• Did size caps bind, or did you set them so wide they never mattered?
• Did stops exist on paper, or only in your head?
• Would you defend this fill in member chat with a dated board screenshot?

Useless questions:

• “How do I make paper look more green for screenshots?”
• “Can I loosen confidence so something happens tonight?”
• “Should I paste live keys to validate the simulator?”

Separate paper logs from notional research track record and locked pick history. Those honesty tools grade research cohorts. Paper grades your automation wiring. Live P&L, if you ever enable it, is a third object with fees and credit costs (credits explainer).

What is the graduation checklist before live keys?

Do not graduate on mood. Graduate on a checklist:

1. Gates are written and unchanged for a meaningful stretch (not rewritten nightly).
2. Paper fills and skips are explainable from those gates.
3. Shortlist size is capped (shortlist limits).
4. Alert noise is under weekly audit, not silenced in anger.
5. You understand pause-at-$0 credits behavior and non-refundable top-ups.
6. Trade-only Binance keys are prepared with withdrawal permission off (Binance trade-only key hygiene).
7. First live size is less heroic than your ego wants.
8. Someone will actually watch the desk during the first live sessions.

If any item fails, stay on paper. Live is a privilege you earn with a boring log, covered also in the broader paper-then-live workflow post when you want a second angle.

How long should paper run before anyone mentions live?

There is no magic day count that fits every desk. There is a evidence standard:

• Enough sessions to include at least one quiet tape and one noisy tape
• Enough skips that you trust the floors
• Enough fills that size and stop behavior showed up
• Zero unresolved “I don’t know why it did that” rows in the log
• Stable gates for long enough that you are not mid-rewrite

Some operators meet that bar in a focused stretch during a Plus trial. Others need longer on monthly Plus with paper only. Both are fine. What is not fine is treating calendar impatience as readiness. If someone says “it has been a week, turn live on,” ask them to explain the last ten skips. Hesitation is a useful filter.

While you wait, keep improving the human loop: UTC snapshot habit, confidence floors, alert audits, and shortlist pruning. Agent quality inherits research quality. It does not create it.

What paper experiments are worth running on purpose?

Useful experiments are narrow:

1. Raise confidence floor for one week; count how many vanity entries disappear.
2. Add a momentum rollover skip; compare skip rate to prior week.
3. Tighten size caps; confirm the cap binds on the loudest day.
4. Drill the kill switch once on purpose in paper (kill-switch drill).
5. Run watchlist-only eligibility if starring discipline is mature.

Useless experiments:

• Randomly lowering every floor to “generate activity”
• Changing five gates at once so you cannot attribute outcomes
• Chasing paper PnL cosmetics for screenshots
• Importing mega-cap strategies into a mid-cap band tool

Document each experiment with a start date and a single success metric. Paper is a lab. Labs need labels.

How should paper mode sit next to the rest of Plus?

Paper is one module inside a wider terminal. Keep the stack ordered:

1. Board literacy and confidence floors
2. Shortlist discipline
3. Alert hygiene
4. Paper agent fidelity
5. Optional live + credits

Skipping to step 5 because the agent tab looks exciting is how desks burn trial weeks. If you are still onboarding, finish phone verify and the first-week plan before deep agent experiments. If you already have Plus habits, paper becomes the place you encode them so they survive nights and travel. Either way, the research columns remain the source of truth. Paper encodes them. Live, if ever, executes them under costs you already understand. Do not reverse that order because a settings toggle is one click away from curiosity.

Paper mode is successful when skips look intentional. Random silence and random chatter are both signs the rule text is unfinished.

Want the live rank 51-200 board, whales, alerts, and paper Trading agent? Start a 14-day Plus trial at app.breakflare.com (Plus is $399/mo or $3990/yr).

Paper graduation table

Frequently asked questions

Is paper mode included with Plus?
Yes. Paper is the default Trading agent posture on Plus.

Does paper mode place real orders?
No. Real orders require intentional live opt-in, trade-only keys, and credits.

When should I turn live on?
Only after the graduation checklist is honestly green. Not after one green paper day.

What if paper and the board disagree?
Debug the rule text against the same day’s ranked rows. Do not invent a new market theory midweek to protect the bot’s feelings.

Does paper spend credits?
No. Credits debit only on live fills at 0.5% of notional.

Can I revert to paper after trying live?
Yes. Treat reversion as a normal safety move, not as embarrassment. Fix gates, then reconsider live.

Related reading

• How live Trading agent credits work (0.5% of fill)
• Introducing Breakflare trading agent
• Trading agent guardrails: stop, size, kill switch
• Your first week on a Breakflare Plus trial