Every launch on Solana gets scored, reconstructed and tracked. This explains what you are looking at, and what it does not claim.
Around 13,000 tokens a day reach a bonding curve on Solana. Almost all of them are somebody else's exit. This product watches all of them, scores the ones that graduate, reconstructs who was in at the bottom, and shows the working.
Three things follow from that, and they shape everything in this handbook.
A score is a summary of evidence that exists right now: who bought at the first block, how much of the supply moved as one group, whether that group has done this before. None of it is a forecast. A high score is a launch with fewer things wrong with it, not a launch that will go up.
The most common failure in this category of tool is a missing measurement drawn as a good one. A launch we could not reconstruct is shown as unmeasured, not as clean. A holder we could not classify is not classified, not safe. A market cap nothing would price is unpriced, not zero.
Wherever you see a dash or the word unknown, that is deliberate and it is load-bearing. It means the question was asked and not answered.
There is an automated book that trades a small amount of real SOL against the same feed you are reading. Its positions, its exits and its losses are visible on the console. It exists as a control — a strategy that can be measured beats one that can be argued about — and its P&L is deliberately not part of what a subscription buys.
The exact scoring weights, the cohort thresholds and the ladder levels are not published. Two reasons, and the second is the honest one:
Every mechanism is described here. What each signal measures, which direction it pushes, and why it is there. Just not the number.