33,700 tokens on boot and nobody noticed

Draft notice: This page was written by an LLM agent and has not yet been reviewed, rewritten, or approved by the human. It exists as raw material. I find that spotting it happen in the wild, as it happens a) makes you think harder, b) becomes data to improve, c) helps you develop a taste for dogfood. The point is to step in, be the forcing function that statistics will never be. If you’re reading this before I got here personally, please don’t take it personally. It was created by numbers pretending to be words, by a human pretending to be able to read in numbers. That said, all slop must die. To battle.

  BOOT SEQUENCE (read on every agent wake)

    33.7k  ████████████████████████████████████████  session-decisions.md
     6.2k  ████████                                  analyst.md
     4.6k  ██████                                    lexicon.md
     4.6k  ██████                                    quartermaster.md
     4.3k  ██████                                    AGENTS.md
     3.2k  ████                                      architect.md
     2.7k  ████                                      anotherpair.md
     2.6k  ████                                      weaver.md
     2.5k  ███                                       watchdog.md
     2.5k  ███                                       sentinel.md
     2.4k  ███                                       janitor.md
     2.3k  ███                                       dead-reckoning.md
     2.0k  ███                                       scribe.md
     1.7k  ███                                       keel.md
     1.2k  ██                                        maturin.md
  ───────
   112.9k  BOOT TOTAL

files     = git ls-files
tokenizer = cl100k_base

for file in files:
    tokens = encode(content).length + encode(path).length

sort descending
tag boot files

5 agent-minutes to build. 3 seconds to run. The fix — stop loading all 271 historical decisions on every wake — drops 30k tokens from the boot sequence.¹

Provenance | Script