Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/0xW1re/solvedocs/llms.txt

Use this file to discover all available pages before exploring further.

Pump.fun is adversarial by design. A significant share of activity on the platform is not organic price discovery — it is coordinated manipulation, bot noise, and deliberate traps set for retail participants. Alpha Leak runs a dedicated layer of adversarial detectors that identify these patterns and ensure the trading system never acts on manufactured signals.

Threat summary

ThreatDetection methodResponse
Bundle / coordinated buyBundleDetector (5s time windows)Token flagged, bundle_confidence ML feature raised
Bot noiseBotDetector (3 behavioral patterns)Wallet capped at alpha 30, is_bot ML feature set
Serial ruggerCreatorRiskScorer (rug rate, lifetime, insiders)creator_risk_score ML feature raised, anti-signal trigger
Copy-trade follower signalCopyTradeDetector (directional consistency)Signal downweighted, originator vs follower resolved
Wash tradingToken risk scorer (buy/sell pattern analysis)wash_trade_pct ML feature, anti-signal trigger at >30%
Exit liquidity trapAntiSignalEmitter (smart wallets selling into retail)Anti-signal emitted, open position force-exited
Crowded tradeSignalCrowdingDetector (tracked SOL vs curve SOL)Score penalty up to 50%, result cached in Redis per token

Bundle detection

Bundles are the most common form of coordinated manipulation on Pump.fun. A group of wallets — often controlled by the same operator — buys a token within the same few seconds, typically at or near launch. This creates the appearance of organic momentum while actually front-running the curve for a later dump. The BundleDetector groups trades into 5-second time buckets and identifies clusters of 3+ wallets buying the same token in the same window. It then scores each cluster on two independent dimensions. Amount uniformity — coefficient of variation (CV) below 0.3 means the buy sizes are suspiciously similar. Random wallets rarely buy the exact same SOL amount. Rank continuity — a rank span at or below the wallet count means the buy ranks are consecutive. Consecutive buy ranks in a tight time window indicates the wallets transacted in rapid succession from the same source.
ClassificationCriteriaStrength
same_slot_coordinatedBoth uniformity and rank continuity passStrongest evidence
similar_amountsUniform amounts, non-consecutive ranksModerate evidence
time_windowOnly time proximity, no other signalsWeakest evidence
The confidence score (0.3–1.0) feeds into the ML model as token_bundle_confidence and into the AntiSignalEmitter as an independent trigger.

Creator risk scoring

Not all creators are equal. The CreatorRiskScorer builds a longitudinal view of every creator with 2+ tokens on-chain. The most useful metric is the rug rate: the fraction of a creator’s tokens whose last trade occurred within 10 minutes of launch and that never graduated. A creator with a 90% rug rate has a consistent and deliberate pattern of launching tokens that die immediately — typically because the creator or insiders dump before any retail participation. Additional signals computed per creator:
SignalDescription
Serial velocityTokens per day over the last 30 days — high velocity combined with high rug rate is an extremely reliable indicator of malicious intent
Avg insider presenceAverage count of insider wallets across the creator’s tokens
Avg bot buyer pctCreators whose tokens attract almost exclusively bots are likely farming interaction metrics
The creator risk score (0–100) is a direct ML feature and an anti-signal trigger at a threshold of 80.

Anti-signal logic

The AntiSignalEmitter is the last line of defense before capital is deployed. It scans every token with an active buy signal every 30 seconds, checking six independently computed risk signals. The multi-trigger design is intentional. Any single metric can be noisy — a 70% bot buyer rate on a brand-new token might just mean no humans have heard about it yet. When 3 or 4 triggers fire simultaneously, the evidence is overwhelming.
Anti-signal fires when ≥ 2 of:

1. Creator risk score > 80
   "Creator [X] has 85% rug rate across 47 tokens"

2. Insider buyer pct > 40%
   "62% of buyers are insiders/bundled wallets (8/13)"

3. Exit liquidity pattern
   "4 smart wallets selling (12.3 SOL) while retail buying (3.1 SOL)"

4. Wash trade pct > 30%
   "41% of volume appears to be wash trading"

5. Bot buyer pct > 60%
   "78% of buyers are bots"

6. Bundle confidence > 70% + bundle buyer pct > 30%
   "Coordinated bundle (82% confidence, 45% of buyers)"
Each anti-signal payload includes the full evidence object — not just that it fired, but why, with the specific numbers. This is logged and stored for auditing.
Anti-signals published to trade:signals are consumed by the live trader, which immediately force-exits any open position in the flagged token. There is no grace period.

Exit liquidity detection

The exit liquidity trigger deserves special attention because it is one of the most reliable indicators of a dump in progress. Detection follows a three-step pattern.
1

Confirm prior smart-wallet buy signal

A token must have had smart-wallet buy signals in the last 15 minutes. This establishes that tracked wallets created the appearance of a good trade and may have drawn in followers.
2

Detect those same wallets selling

The same smart wallets that generated the buy signal are now selling — within the last 10 minutes. This is the reversal that makes the setup dangerous.
3

Confirm retail is buying in simultaneously

Retail wallets not in the tracked set are actively buying while the smart wallets are selling. This is the classic exit liquidity structure: smart money is using retail inflow to exit the position.
The detector computes the tracked sell SOL vs retail buy SOL ratio as part of the evidence, which helps quantify how significant the exit is relative to new inflows. A ratio above 1.0 — where tracked wallets are selling more than retail is buying — is a near-certain indication of a controlled unwind.

Build docs developers (and LLMs) love