~/reversepad > ./lab

four simulations of what reversepad does, on real data: saved reads of real coins, the real name rule, the real launch transaction, the real pattern library. nothing here costs anything or sends anything.

01 reverse replay

pick a coin. this replays a real read the helper made: what it looked up, which files it read, every line it decompiled with the source it came from, and the three forks it wrote. sped up, nothing re-run.

  1. 00lookup
  2. 01capture
  3. 02evidence
  4. 03decompile
  5. 04forks
files read0
kilobytes0
lines0
with a source0
unknown0
A website being read: page, scripts and styles
evidence
    dna.card
    02 name guardthe helper's own rule

    a fork can never ride the source's name. the helper refuses any fork name or ticker that matches the source, sits within 2 letter edits of it, contains it, or has a word too close to it. type a name and watch the edit distance get worked out.

      edit distance ·
      03 will it fit?solana limit: 1232 bytes

      a launch is one solana transaction: create the coin, and buy the first tokens in the same approval. it has to fit in 1232 bytes. this builds the exact transaction the launch page would send, with the same code, for a throwaway wallet. nobody signs it and it is never sent.

      first buy
      press build it. the launch code (about 1 MB) loads the first time.
      04 pattern graph

      every dna card names the patterns a coin uses. these are all of them, from the studies behind reversepad: the inner ring is the patterns, grouped by kind; the outer ring is the projects they were seen in. touch one to see its links. the coin picked in the replay lights up the patterns its card used.