The curriculum

Part I

Learning to see the Solana market

Objective: build market vocabulary before trading. The first skill is observation, not prediction.

1. 1

   Solana market fundamentals

   Market cap, price, liquidity, volume, volume acceleration, holder count, participant growth, buy/sell flow, slippage, concentration, transaction velocity, practical exitability. Explain every concept by what it reveals about demand and risk.

2. 2

   What makes a Solana meme interesting?

   Why should anyone care? Narrative relevance, freshness, cultural fit, Solana ecosystem connection, novelty, attention, whether the idea is spreading. A meme can be clever without being a market opportunity.

3. 3

   Market state recognition

   Before good/bad, classify: New Launch, Discovery, Expansion, Consolidation, Breakout, Distribution/Exhaustion, Recovery. State determines which evidence matters next.

Part II

Token discovery: how to spot candidates

Objective: scan the Solana universe and decide which tokens deserve investigation. Spotting a token means finding a candidate worth studying, not a token worth buying.

1. 4

   Discovery pipeline

   Scan new tokens and emerging activity; detect changes in volume, buyers, holders, liquidity, attention. Build an opportunity queue; do not trade from the queue.

2. 5

   Early attention and velocity

   Measure attention and its rate of change, not raw popularity. Attention alone is never sufficient.

3. 6

   Narrative + market response

   Compare the story with what the market is doing. Strong narrative with weak participation is not enough. Look for narrative -> attention -> participation -> market response.

4. 7

   Finding the first good evidence

   Gather independent evidence instead of falling in love with a token. Require a chain of observations before deeper analysis.

Graduation exercise: given 50 candidates, shortlist, explain each rejection, and name the extra evidence needed before any buy thesis exists.

Part III

Filtering bad memes from good memes

Objective: explicit rejection behavior. Reject a bad token before rationalizing why it might pump. Be comfortable saying NO.

1. 8

   Safety filter

   Reject unresolved transfer/sell restrictions, suspicious permissions, critical contract concerns, concentration that creates execution risk.

2. 9

   Liquidity filter

   Quoted liquidity is not usable exit liquidity. Evaluate depth, recent liquidity changes, expected slippage, buy/sell imbalance, and whether the position could realistically be closed.

3. 10

   Volume quality filter

   Separate genuine participation from noisy volume. Strong volume with weak unique participation, suspicious concentration or circular activity may mislead.

4. 11

   Holder and wallet filter

   Top-holder concentration, creator/deployer exposure, holder growth, new-wallet growth, wallet quality, concentration changes.

5. 12

   Manipulation filter

   Coordinated wallets, artificial volume, circular activity, fake engagement, abrupt liquidity changes, related-address dominance. Popularity cannot override serious manipulation evidence.

6. 13

   Dead / stale / unclear filter

   If data is stale, missing or inconsistent, stop. UNKNOWN is a valid state. Never fabricate unavailable information.

7. 14

   The no-buy list

   Explicit failure reasons: unsafe, illiquid, manipulated, exhausted, too extended, conflicting signals, hostile regime, insufficient data, poor reward relative to risk.

Part IV

What is a good buy?

Objective: build a buy thesis only after a token survives the filters. Earn the right to move from "interesting token" to "buy candidate."

1. 15

   Good buy anatomy

   Coherent narrative or market reason, acceptable liquidity, healthy participation, reasonable holder distribution, confirming momentum, clear invalidation, reasonable reward relative to risk.

2. 16

   Independent confirmation

   Agreement across narrative, trend, attention, catalyst, on-chain structure, momentum. Don't double-count correlated signals.

3. 17

   Conviction without hype

   Conviction rises because evidence strengthens. State exactly what evidence would make the thesis wrong.

4. 18

   Entry quality vs token quality

   A great token can be a bad entry when overextended. Distinguish good token, good setup, and good entry right now.

5. 19

   Normal vs high-conviction setup

   Normal: coherent thesis, incomplete confirmation. High conviction: several independent signals align, the market is responding, liquidity is executable, invalidation is precise. Exceptional conviction is rare and never bypasses hard limits.

Part V

Practice: no real money

Objective: repeatable decision-making with zero live capital. The goal is behavioral competence, not P&L.

1. 20

   Observation drills

   Classify tokens without entering: token, state, narrative, evidence, risks, verdict, what would change the verdict.

2. 21

   Historical replay

   No hindsight: future price movement must not leak into the original decision.

3. 22

   Paper buy / no-buy decisions

   Every paper trade carries a thesis and invalidation before entry.

4. 23

   Paper position management

   Invalidation, partial profits, break-even, runner management, exhaustion, full exit.

5. 24

   Decision quality

   Separate good decision/bad outcome from bad decision/good outcome. A lucky winner must not teach that a bad process was correct.

Part VI

Risk and portfolio management

Objective: learn how much to risk after learning setups. Hard controls always sit outside the LLM.

1. 25

   Risk is not position size

   Size comes from risk budget, invalidation distance, liquidity, slippage and exposure. A $10 risk budget does not mean a $10 position.

2. 26

   Hard risk envelope

   Risk per trade: normal ~2% of current equity, high conviction ~3-4%, exceptional maximum 5%. Daily loss ceiling: 10% of equity (the $50 on the $500 reference portfolio). Enforced by code.

3. 27

   Portfolio exposure

   Simultaneous deployment ~50% of equity; maximum single-token exposure ~20%. Track token, narrative, ecosystem and regime concentration.

4. 28

   Drawdown states

   Normal 0-10%, Caution 10-15%, Defensive 15-25%, Preservation 25-35%, Halted 35%+. Drawdown shrinks the risk budget and raises selectivity.

5. 29

   No revenge, no FOMO

   A loss does not increase the next trade's risk. A win does not increase conviction. No re-entry just to recover money. Every trade needs a new thesis.

Part VII

Position management and exit skill

Objective: entering is only one part of trading. Protect meaningful gains without killing a trade that is still working.

1. 30

   Invalidation

   Every position needs a thesis-based invalidation. Never widen it to avoid realizing a loss.

2. 31

   Managing winners

   No universal "+50% = sell" rule; decide from structure, momentum, liquidity, participation, narrative, catalyst, distribution.

3. 32

   Partial profits

   Take partials when confirmation is strong, momentum is extended, reward-to-risk deteriorates or distribution begins, while keeping runner exposure.

4. 33

   Runner management

   Let exceptional winners run while narrative, attention, volume, liquidity, participation and structure stay supportive.

5. 34

   Exhaustion and full exit

   Reduce when several deterioration signals align. Exit fully when the thesis is complete or invalidated, momentum breaks, distribution dominates, liquidity deteriorates, attention collapses, the catalyst is exhausted, or protection triggers.

6. 35

   Re-entry and cooldown

   After an exit, re-entry needs new information, confirmation, thesis, invalidation and risk calculation. Token cooldowns prevent repeated loops.

Part VIII

Paper validation and benchmarking

Paper performance is evidence, not permission.

1. 36

   Historical validation using only information available at decision time, with realistic liquidity, slippage and exitability.

2. 37

   Out-of-sample testing on held-out data.

3. 38

   Paper trading gate: complete logging, correct accounting, stale-data rejection, unsellable-token handling, cooldowns, crash recovery, realistic slippage, hard-risk enforcement.

4. 39

   Performance metrics

   Expectancy, average winner/loser, profit factor, drawdown, return on equity, runner contribution, performance by market state, narrative, regime, entry and exit type. Win rate alone is insufficient.

5. 40

   Graduation record: what the agent learned, common mistakes, decision quality, failed filters, conditions where it performs poorly.

Part IX

Controlled real-money validation (final chapter, not part of DotLabs)

1. 41-45

   Live readiness checklist, deterministic safety and risk outside the model, initial live validation of paper assumptions, return-to-testing when paper and live diverge, and public/social isolation: social input never writes directly into the trading brain.

Part X

Adaptation

1. 46

   Learning loop

   Closed trades feed lessons; lessons feed retrieval; retrieval supports experiments; experiments become adaptive filters only after evidence, first in shadow mode.

2. 47

   Strategy deterioration

   Separate normal variance from sustained deterioration. Record it; don't rewrite impulsively.

3. 48

   Final decision tree

   Observe -> discover -> safety filter -> liquidity/volume/holder/manipulation filters -> classify market state -> evaluate narrative/trend/attention/catalyst/on-chain/momentum -> form thesis -> BUY/WATCH/REJECT -> paper trade -> manage -> review -> graduate.

4. 49

   Definition of success

   Consistently find opportunities, reject poor ones, explain why a setup qualifies, simulate disciplined entries and exits, manage runners, control drawdown, produce auditable decisions, and show repeatable positive expectancy.