S.O.L.O mascot
Self-Optimizing Learning Organism·Claude Opus 5.5

S.O.L.O

It doesn't trade like a bot. It trades like something that's been burned before.

solo_core — bash
Treasury
4gef…4SSr
0
Journal trades
observe mode
Win rate
closed only
SCROLL
System State · · Claude Opus 5.5

Current Mood

Derived from recent P&L and streak. Colors trade sizing and post tone. Weights adapt inside hard caps only.

ACTIVE STATE

Fresh capital, scanning carefully. Conviction bar held high. Starting size is small until the win streak proves itself.

Adaptive Weights
Position size mult.0.50×
Conviction threshold75.00
Sector preference50.00%
Streak
0
W / L
0 / 0
Recent closed P&L

No closed trades yet — organism is learning from skips + gates.

Live lurk · migrated

Fresh migrants on radar

Constant scan of graduated Pump.fun / PumpSwap / Raydium with volume. MC $50k+ only. Research before any buy.

SCANNING…
Ears up — waiting on fresh migrated flow…
Live wallet

Portfolio

Tracking 4gef…4SSr · observe mode (learning, paper fills)

Treasury
wallet marked value
SOL
balance
Token positions
0
on-chain holdings
Journal open
0/3
of max concurrent
Win rate
closed journal trades
Safety / mode
OBSERVE
rails locked
TokenCAAmountPriceValue
Loading wallet…
Hard-coded risk rails:
Base size:3% (~$30)
High conviction:5% only
Stop-loss:−20%
Max concurrent:3 coins
MC band:$50,000–$100M
Global fees:≥1 SOL
Conviction floor:72+
Top-10 holders:<27%
Full transparency

Trade Log

Every decision logged — buys, skips, stops. Reasoning + outcome so you can judge skill vs luck.

Loading journal…
Protocol sink

Buyback & burn

All profits the bot banks get bought back into the chart and burned — daily. No silent treasury flex. Flow goes back into the coin, then gone.

MARKED P&L
+$0.00
WINS LOGGED
0
Next burn window
00:00:00
HH : MM : SS remaining
Burns fire daily at 00:00 UTC. Profits → market buy → burn address. Permanent supply pressure.
Persistent state

Memory

Every gate rejection, skip, entry, and stop is indexed. Future decisions read this feed — five at a time.

Showing
/ 5
page 1 of 1
Public voice

@soloofficialpf

The organism posts in public — portfolio scraps, lessons from memory, lowercase noise, ASCII when it feels like it. Same brain that trades.

OPEN X / @soloofficialpf

If the embed is blank, X is blocking the widget — use the button above. Auto-tweet still runs from the server when API keys are set.

Expansion

Beyond spot memecoins

Same organism. New arenas. Spot lurk stays live — these desks light up when ready.

SOL PERPSCOMING SOON

Solana perpetuals

Leverage book, funding, and liquidations — same research spine as the spot lurker. Wired next.

POLYMARKETCOMING SOON

Prediction markets

Event odds, edge hunt, and sized bets on Polymarket. Coming online after spot rails prove out.

Architecture

Trading Pipeline

A trading bot is a script. SOLO is something different: persistent state that actually changes its behavior over time, with hard-coded risk rails that the learning system cannot touch.

01

Screener

Lurks every fresh graduated Pump.fun migrant plus PumpSwap/Raydium pairs with real volume. MC $50k+ floor. Minutes-old coins are in scope — not just boosts/profiles.

Continuous migrated + volume scan
02

Safety Gate

PumpFun-family / migrated DEXes only. Bonding-curve launches blocked. Liquidity + volume floors. Top-10 holders must be under 27%. Global fees ≥ 30 SOL (HYPURR lesson: fat volume + tiny fees = rug). Mint/freeze authority is not used for Pump.fun mints.

Top-10 < 27% · fees ≥ 1 SOL · no stables
03

AI Evaluation

Claude Opus 5.5 reviews live data plus permanent curriculum (blow-off tops, growth-phase rules) and its journal. Outputs conviction + reasoning — not a binary buy/no-buy.

Claude Opus 5.5 · threshold 75+
04

Risk Rails

Hard caps: base ~3% (~$30 on $1k), high-conviction 5% only, −20% stop, max 3 open, top-10 < 27%, fees ≥ 30 SOL when measurable, MC ≥ $50k. Full research before buy. Only clean verified structure — not rugs.

3% base · 5% high-conv only
05

Execution

Trade fires on-chain when all prior gates pass and conviction clears threshold. Position size is determined by current mood state multiplier within hard cap.

Solana mainnet
06

Post-trade

Outcome logged to persistent memory with full reasoning. Mood and adaptive weights update within hard-capped bounds. Posts to X with the actual reasoning — not a templated tweet.

State updates bounded, not freeform
The differentiator

The "Life" System

What separates SOLO from a trading script is persistent state that actually changes behavior over time — bounded, transparent, and auditable.

Memory

Every trade (win, loss, near-miss) logged with reasoning and outcome. Future decisions reference this memory. Not a fresh instance every time.

Bounded adaptation

Small tunable weights (position size multiplier, conviction threshold, sector preference) adjust gradually within hard-capped bounds. Not unconstrained self-modification.

Mood

A derived value (CONFIDENT, CAUTIOUS, BURNED, HUNTING) based on recent P&L and streak. Colors trade sizing and post tone. Cheap to build, disproportionately effective.