42-case first run: 39 were one uncompilable TU (ov_SC04_018_jr_8017AE2C.c), not draft defects —
the resolver's 'undeclared' classification needs a TU-alone compile probe first (open follow-up).
R30/R31 capture while hot: the resolver pivot (63 zero-token banks of 245 staged of 424 judged of
1,352 nominated), the RED-fleet finding (15/214 baseline-RED refusing 174/182 doubly-verified
drafts), the three byte-proven repairs so far, and rule candidates R56–R58.
frontier-analysis-s60 §4 measured that ~571 open functions had FINISHED drafting (closeness-0 backlog
rows / reloc shape-MATCH rejects) and were being re-drafted wave after wave. tools/integration_resolver.py
treats those ledgers as an index: still-open? -> rtu_match at the real split TU (CC1: the gate ladder's
draft-side transforms, one retry) -> reloc_identity as the disagreeing oracle (rtu masks reloc fields)
-> aprop_symfix on MISMATCH/shape-MATCH -> stage -> sweep_parallel (whole-binary SHA, sole arbiter)
-> commit at once (R42). Refuses main by name (gate_main owns it), //@EDIT drafts, dirty trees, collapsed
registries; every drop is counted (R32); a negative control over recently-banked functions must pass
N/N before any verdict is trusted (R35/R39 — its first form picked carve moves as banks, 9/12 FAIL,
and was fixed before a single stock verdict was read). Ledger .run/resolver/verdicts.jsonl keyed by
(binary, fn, draft-sha, split-TU-sha) so unchanged rejects are never re-judged.
First pass (commit:2991): 1,352 nominated -> 901 already banked, 27 main -> 424 judged in 41 s ->
245 staged (57.8%; 242 raw, 3 via transforms) -> 63 banked (net INCLUDE_ASM delta; that commit's
subject says 72 = gross incl. 9 carve moves), 182 gate-refused, zero model tokens, ~10 min total.
Lane wrapper tools/lanes/resolver_lane.sh (holds .run/auto/draw.lock for judge+gate: rtu reads the
TUs a gate splices into).
The audit's headline, measured: THE WALL IS AN INTEGRATION WALL, NOT A CODEGEN WALL. Of the 292
functions the gate has refused 6+ times, 178 (61%) have ALREADY produced a closeness-0 draft —
match_one byte-equality, whole-binary gate rejection. The blocker is symbols/decls/TU plumbing,
and the fleet keeps re-drafting them: 10,049 reject rows over 574 distinct functions. Highest-EV
build is a zero-token integration-resolver lane, not more drafting.
CORRECTIONS TO MY OWN NUMBERS, verified against the tree before accepting:
* siblings are 1,334 behind 480 multi-member groups, NOT ~3,900. 1,292 groups are SINGLETONS
carrying 57% of open instruction mass. I conflated the never-drafted stub count with the sibling
count and overstated remap leverage ~3x, in this checkpoint and repeatedly in conversation.
* 'everything drawable is gen6+' holds only for the collapsed wave-eligible view; whole-pool
generation is 53% gen0/1, 25% gen6+, and only 292 fns are 6+ GATE-refused.
* '30-67 min gates at 8% CPU' conflated wall_min (includes drafting/queue) with gate wall (12-31
min healthy). Gate cost is proportional to FAILURES, not drafts: ~3 whole-binary builds per
failing draft, so banks/gate-min fell 17.5 -> 0.10 as conversion fell.
* the 5,388 closeness<=2 rows de-dupe to ~543 open functions; my own 19:40 re-measure found 290
still open, down from its 470 — the re-gate and grinder are draining that pool now.
* campaign_status's 'banked today' undercounts: the stub invariant says ~2,644 net, because the
A-prop lane's 357 rode in a chore commit its regex cannot see.
One documented counterexample to 'model quality is not a bottleneck': func_80181714, where
ox-alpha plateaued at closeness 4 while Opus/GLM/DeepSeek each reached reloc-verified MATCH —
argues for a small escalation tier AFTER the resolver drains the fake walls.
Taken on trust and flagged as such: the A-prop residual split (169 STRUCT / 121 no-seed-decl /
73 IMM / 12 void) — the refusal mechanisms exist in aprop_autodraft.py but no file carries those
counts; re-derive before building the decl-inference tool.
Five reviewers on disjoint wave groups (D1 dd/de/df · D2 dg-dm · D3 cg-cm · D4 cn-cw ·
D5 cx-dt): 859 COVERED · 46 ADDENDUM · 9 NEW · 289 REJECT. Index now 888 sections.
§283 the 46 sharpenings, one block per target section
§284 combine can reassociate two sequential bitwise-AND masks against the PRE-mask value;
an asm fence at the mask's definition point stops it
§285 pre-initializing a variable with a shared constant BEFORE a branch makes both arms of
the following if/else destructively reuse one register
§286 fold a statement's side effect into a comma-expression in an argument position to
place its RTL relative to a call's own delay slot
§287 a stack-frame hole below two address-taken aggregate locals is a LEADING PAD MEMBER of
one combined struct, not separate locals
§288 a register pin declared UNINITIALIZED and assigned only at its late sole use still
forces the fixed register's save/restore
§289 an array local's decayed base keeps EVERY element's store alive under DSE, though only
one pointer value escapes
§290 one strength-reduced giv can drive stores to several distinct relocatable symbols,
each keeping its own %hi/%lo anchor
§291 the delay-slot false-value: a conditional branch's zero arm must be a fall-through
adjacent block ending in an explicit goto
§292 declaring a symbol upstream of an already-banked sibling that relies on its implicit
(K&R) declaration silently reprototypes the sibling's call site
VERIFICATION CAUGHT SIX BAD CLAIMS, recorded in §283 as refuted rather than laundered in:
three independent "the harvester leaked an unbanked NEAR into a banked-only harvest" reports
(all three functions are genuinely banked per corpus.stubs — the notes predate their gate and
read as harvester bugs hours later); a "match_one resolves targets by bare symbol name" claim
(it resolves an explicit path, and api_draft always passes it — the real hazard is its
resident-defaulting --asm-subdir, hardened separately in commit:2917); and two idiom claims
whose mechanism is absent from the banked code. §284's fence was described as NON-volatile
and the banked code uses `__asm__ volatile` — corrected in the text, since that is a detail
readers copy verbatim.
THE STRONGEST SIGNAL IS NOT A SECTION: eight cards across four waves independently
re-derived that the whole-object gate needs every sibling matched. It is implicit in the
corpus and has never been stated as its own law. Recorded in §283 as the batch's clearest
missing-section signal.
PROCESS, for the next batch: forked sub-reviewers exceeded their brief in three of five
groups — one re-derived five waves it was not assigned and self-merged over the shared output
path, one silently dropped 11 rows including a whole function, one produced nothing. Each
parent caught its own fork. Tell forks not to spawn forks (they infer it from the parent's
inherited context) and give each a private output path.
34 candidates: 13 COVERED · 8 ADDENDUM · 4 NEW · 9 REJECT.
§278 the 8 sharpenings (§17, §20, §74, §162p, §164-56, §172b-4, §237, §265)
§279 a do/while (p < end) under an entry guard — the GUARD decides the compare
§280 the cursor's declared type picks sltu vs slt: a `T*` bound test is unsigned by
C rule, an integer cursor is signed — confirmed twice in one wave by unrelated
functions, which is why it generalizes to the corpus's scan-loop residuals
§281 group copy-then-RMW by OPERATION KIND, not field by field; sched1 does the
interleaving. §213 explicitly disclaimed this shared-pseudo case — a named gap now filled
§282 gcc's own loop reversal places the reversed counter's init AFTER the hoisted
movables, a position no hand-written down-count loop can reach — resolves a class of
"prologue order" residuals that looked unfixable from the C side
VERIFICATION, both directions (R14/G3). Two reviewer claims were REFUTED at byte-check and
recorded as refuted rather than laundered in: func_800CAE74's note described a "parameter
reassignment" idiom absent from the banked C (the real law became §280), and func_800CBCD4's
named the wrong register ($v1 where the bytes use $v0). I re-verified §280 independently at
merge: the target carries one `slt` and zero `sltu`, and the banked C drives the bound test
with a plain s32 cursor beside a separate s32* — the direction the law predicts.
Section titles were rewritten at merge: the reviewer put each law in the body and left the
heading as "(func_X, ov_Y)", which is unfindable. A section is written to be GREPPED, so the
title states the law.
Three harness suspicions recorded in §278, NOT acted on — func_8017F39C suspects a
binary-registration mixup with a same-named function in ov_SC01_009, which is a gate-side
question, not a drafting one.
Index: 824 sections. Distill state updated, cf marker consumed; the lane is free to raise the
next batch.
The ore that piled up behind a stale READY marker for 10.5 h (waves be..ce). Three Sonnet
reviewers on disjoint wave groups read all 315 harvested notes against the corpus:
255 COVERED · 21 ADDENDUM · 3 NEW · 26 REJECT
255 already-covered is the flywheel working, not waste — §193-§273 are earlier rounds of
this same cycle, so a note restating a law we own is the corpus doing its job.
§274 the 21 sharpenings, one block per target section (§8, §20, §82, §134, §164-51,
§164-64, §172a, §172b-1, §172b-4, §176-F5, §179-C, §179-D, §211, §219+§1-I5,
§224, §225, §225/§256, §226, §229, §244, §136d-1)
§275 THE LEFTOVER-REGISTER READ — an entry block that consumes $v0/$v1 with no earlier
write in the function is reading the CALLER's return value; reproduce with an
unassigned `register s32 v __asm__("$2")`, and stop hunting for a dropped call
§276 MIXED ADDRESS-EXPRESSION SPELLING as a CSE-unification dial — spell one occurrence
of a repeated symbol as offset arithmetic on its neighbour ((&D_X)[1]) and cse
cannot unify them; found twice, independently, on the same function
§277 RETURN-TAIL C SPELLING picks the delay-slot-fill vs trailing-move topology
VERIFIED AT MERGE, not taken on trust (R14/G3): the leftover-register read, the CSE-defeat
spelling, the compound-rounding pair (live in ov_SC03_095), and func_801846F0's delay-slot
`addu $v0,$zero,$zero`. Two reviewer claims did NOT survive and are corrected in place:
§275's law was stated as "no call anywhere in the body" when the function has five jal's —
what makes it true is that no write to the register PRECEDES the read; and a claimed
harness defect (func_800CE004 "will not bank") is false, it has a real body at
src/md_MAIN_044/md_MAIN_044.c:443 and corpus.stubs does not list it as open.
B and C reached the same CSE lever from different waves and the §1-I5 / §219 rounding law
arrived twice on three overlays — both unified into one entry rather than duplicated.
Index regenerated: 806 sections. The lane immediately raised its next batch (cf, 34
candidates), which is the proof it is unblocked.
.run/maintenance.sh (what runs) and tools/lanes/maintenance.sh (a pre-S59 copy) had
diverged. The 150->50 threshold tune landed on the stale copy and was then copied over
the live one, silently reverting five S59 fixes:
* the R47 shape filter (staging fell back to status=='AGREE' alone — the exact defect
that staged 82 hopeless drafts every 45 minutes)
* the R48 (binary, fn) keying (bare-fn keys collide across overlays)
* reloc --fix MISMATCH auto-repair (measured 4/4 repaired to AGREE)
* rtu_second_chance (re-judges standalone COMPILE-FAILs against the real TU)
* fix_tu_ret_decls (the return-type half of the stale-decl wall)
Rebuilt from the S59 lineage with the 150->50 threshold and the periodic fleet R22
re-applied, both paths now byte-identical, `bash -n` clean, and the two-path hazard
documented in the header so the next edit cannot repeat it.
Also: relaunch_drafter_shell.sh 30s -> 5s ready-marker poll; regenerated backlog and
fleet progress artifacts.
docs/tool-designs/aprop-lane-s59.md: baseline (117 staged / 0 banked = 82
shape-DIFF + 19 standalone-fail + 16 near-0, none invisible), the defect
classes with byte-proofs, the fixed pipeline's end-to-end numbers (64+ banked,
zero tokens), the near-miss taxonomy (11 IMM tier-2, 8 wrong-family), the
ceiling with every refusal named and sized, and the two pre-existing RED
binaries the scoped R22 surfaced (SC07_010 fixed green; SC07_002 named for the
jtbl owner).
Cookbook: §270 the four-verdict law (instructions/symbols/TU/whole-binary —
stage on fewer and burn a build per missing layer forever), §271 ordinal
candidate pairing, §272 K&R the definition, §273 the wrong-oracle law.
maintenance.sh now runs the full recovered pipeline unattended: shape-gated
(binary,fn)-keyed staging, reloc --fix repair+recheck, rtu second chance,
fix_tu_ret_decls on fresh near-0 rejects.
PROVEN: from 14:57:01 to 18:43:23 today HEAD built main to 307aa45d… against the
expected 143dbb89…, with NO draft substituted (measured under gate.main.lock, no
gate_main alive). Auto-commit commit:2693 had adopted a mid-flight gate_main
substitution — its carve-out reverted main's TUs, gate_main re-wrote them, and
`git add -A src/` swept the unverified bodies in (a TOCTOU race, 14 s after a
bisect chunk banked). Every main batch after it was doomed before its first
draft was judged: m00–m03 card cycles drafted ~737, slated 160, banked 0, and
burned ~50 clean rebuilds bisecting innocent slates. commit:2712 restored the
green content by accident (it swept this investigation's diagnostic checkout).
gate_main: on any batch failure, ONE try_batch([]) control runs first — if HEAD
itself is red it prints BASELINE RED, leaves the slate reusable, exits 3 (R40).
clean_build no longer reports a linked-but-mismatched build as "no binary" (the
build target embeds the SHA check), the compile-conflict shortcut fires only on
error-shaped lines naming a symbol some draft in the slate actually uses (the
baseline's own func_800143AC implicit-decl WARNING was matching — every m04
chunk died with "drafts declaring it: []"), reverts narrow to top-level src/*.c
(main_tus) so a main gate can never destroy overlay lanes' in-flight work, and
--assert-baseline is a first-class mode.
main_lane: every cycle opens with gate_main --assert-baseline and REFUSES to
draft or gate against a red baseline (R43) — BaselineRed parks nothing, burns
no tries, writes .run/main_lane.BASELINE_RED, re-checks every 30 min.
Adopters (ox_campaign ×3, maintenance.sh, gate_stage, gate_lane, idiom_serial):
main's TUs (top-level src/*.c) are never staged and never reverted by an
overlay/maintenance lane — one writer (gate_main), one committer (main_lane,
after the whole-EXE SHA re-checks green). Unstage-after-add is race-free where
the old revert-then-add was the losing half of the TOCTOU.
Diagnosis, evidence and the full timeline: docs/tool-designs/main-lane-fix-s59.md
I reported 240 of 244 turn-finishes truncated in wave bk and called it ~100%. That
count came from grepping lines containing 'finish=', which api_agent only prints when
a turn ends WITHOUT a tool call — so the denominator was not all turns, it was all
anomalous turns. I compared a subset against itself.
Counting every turn:
bk (8k / 420s): 240 truncated of 3,222 turns = 7.4%
bt (16k / 700s): 16 truncated of 1,210 turns = 1.3%
Still a ~6x improvement and the change stands on its own evidence, but it is a tax
reduction, not the collapse I described. Recorded alongside: a truncated turn is one
turn of 24, not a lost agent — the logs show the agent emitting its tool call on the
very next turn.
Both docs keep the wrong figure explicitly, with why it was wrong, so the next reader
does not re-derive it from the same grep.
SETUP gains rows for recover_rejects.py (free recovery of the 45% of drafts that never
reach the gate, 13% of which are a deterministic symbol rebase),
restart_main_lane_when_idle.sh, and a campaign-constants row recording MAXTOK 16000 /
HTTP_TIMEOUT 700 with the measurements behind them: reasoning_tokens=0 so the output
cap was the reasoning cap, 240 of 244 turn-finishes truncated at 8k, an uncapped hard
prompt wanting 8,067 tokens, ~30 tok/s, and the 1M-context / 131,072-max-completion
model ceiling that makes 16k our choice rather than a limit.
The runbook gains the same table plus the ordering rule (generation < HTTP_TIMEOUT <
stallguard's 1200s kill), the evidence that turn caps are NOT binding on the default
lane, one-lane-one-band with the size table that retired the 120-2000 slot, the
maintenance lane's new recovery job, and the main-lane restart helper.
Drew's endgame deliverable is a workflow system another person can run solo on any
target, any compiler. We had three docs recording BFM's history (accelerators,
decision-log, automation-runbook) and none stating the SYSTEM. This is that document,
written to one rule: if it would still be true for a different console and a different
compiler, it belongs here.
Contents: the byte gate as the only arbiter and the oracle ladder around it (each
oracle must state its own blind spots); lanes and why the clock-limited one is never
stopped to ship a change; the CARD as fuel rather than a ticket, including the lever
lesson (a label a worker cannot look up is a dead end — 108 transcripts searched for a
word our knowledge base did not contain); draw-time refusal and the four measured
instances of the dominant defect class; free work before paid work; giving an excluded
population its own lane instead of an exclusion; model routing and per-lane budgets;
the flywheel with the one-wave lag and the inert-rider law (1 in 3 credited levers is
byte-inert, measured twice); the economics as measured; what transfers vs what does
not (the knowledge base is per-compiler, the machine that builds it is universal); and
a bootstrapping order for a new project.
Numbers are stated with their denominators so a new project calibrates rather than
copies.
The runbook was dated 2026-06-22 and described the reach-1 grinder pivot — it named
no lane that exists today, two months and an entire toolchain later. Rewritten around
what is actually running: the six lanes and their restart rules, the OpenRouter
drafting toolchain (cards, LEVER_CRIB, per-lane budgets, the draw's admit/refuse
census, quotas as floor AND ceiling), the banking toolchain by binary class
(sweep_parallel vs gate_main vs the gate-time jtbl carve vs -O0 objects), the main
lane, the distill lane and the flywheel's measured yield, the rate/credit numbers with
their denominators, and recovery.
Two ops laws are stated where they will be read rather than rediscovered: bash parses
a while-loop up front (so code, args and draw-defaults each take effect differently),
and never pkill -f a lane by a bare name because it matches the harness's own wrapper.
SETUP.md gains rows for main_lane, the distill lane and the three restart helpers.
268 a register __asm__ pin on a call-clobbered register is honored EXACTLY when the
pinned range crosses no call — turning 257-2's 'pins are silently ignored' into a
rule with a precondition, and explaining both faces: a pin that does nothing
(range crosses a jal, gcc silently falls back) and a pin that fixes a REGALLOC
residual in a function that has calls (range sits between them). Three A/B'd cards.
269 ten addenda, plus the 266 solo-lever sweep table.
THE HEADLINE IS THE INERT-RIDER RATE. 19 strip-tests across 13 banked bodies: 6 of 19
credited levers are byte-inert (32%), against 4 of 8 last batch. One in three 'this is
what made it match' claims credits something that changes nothing — and one whole
proposed section dissolved under its own strip test. 67 of 82 candidates (81.7%) were
already covered, matching the previous batch's re-derivation rate.
Two discovery-gap defects the distill review measured, both costing drafters real
compiles:
1. LINE NUMBERS CITED AS SECTIONS. Index rows end with a <sub>L1234</sub> anchor = the
section's line in the cookbook. Drafters read it as a section id and cite it: this
batch alone carries §1907, §12479, §2965, §11383, §8892, §5583, §1832, §2429,
§1755, §2609 — line numbers, every one, and a grep for any of them returns nothing.
The number is real and the reading is wrong, which is the worst kind of dead end
because it looks like a citation. The index now publishes an L→§ table, and
answers it directly (verified: §12479 → §3-The,
§1907 → §21).
2. THE INDEX WAS NEVER NAMED AS THE ENTRY POINT. One drafter wrote 'no numbered
section I could find by grep' about a lever whose section title literally contains
the words it searched — it grepped the 25k-line cookbook, not the symptom-keyed
index. api_agent's SYS now says to start at the index, and how to resolve a
line-number citation.
265 the verbatim-asm bank lane — two in-tree precedents, the MASPSX decimal-immediate
rule, and a REVIEW-ADDED accounting caveat: an __asm__ body is not an INCLUDE_ASM
stub, so corpus.stubs() counts it as MATCHED. Bytes proven, function not
decompiled. Fine for hand-written asm; for an -O0 C function the right answer is
the -O0 object, and this lane is a temporary hold at best.
266 the inert-rider law — a lever is only citable when its solo removal breaks the
match. Measured 4 of 8 credited levers on this batch were byte-inert: a $2 pin
silently dropped, a volatile the scheduler already ordered, a named zero, and a
statement split. The banked artifact CONTAINS the rider precisely because it is
inert, which reads as proof. R40 applied to the flywheel itself.
267 eleven addenda to existing sections, six with fresh match_one A/Bs.
The batch's real headline: 145 of 165 candidates were ALREADY COVERED, and 8 claims
were refuted (4 by live A/B) — the harvest is mostly re-derivation, and an unchecked
wrong law is worse than no law.
Three measured harness defects, all fixed:
1. THE CARD NAMED A WORD THE COOKBOOK DOES NOT CONTAIN. api_agent stated the lever
as a bare label and nothing else; grep 'extend-tell' / 'swaprepeat' / 's16-div-tell'
over the 750-section cookbook returns ZERO. 108 failure transcripts grepped
extend-tell and 28 grepped swaprepeat against nothing while the knowledge sat at
172a/172b under different words. Fixed both ends: a LANE ALIASES grep-bait block at
172b, and LEVER_CRIB on the card — what the tell means, the section to grep, and
the byte-proven C spellings.
2. ONE GLOBAL AGENT BUDGET FOR CARDS OF VERY DIFFERENT SIZE. tells cards are 2.4x the
default lane's (median 89-95 ins vs 37-39) and stack 3-5 idioms; 98 of 270 final
attempts ended AT the 24-turn cap. LANE_BUDGET gives tells 40 turns / /bin/bash.40, and
logs the choice so it is auditable rather than invisible.
3. Two new SYS laws: grep the section your crib cites before drafting, and stop when
the residual class says [permuter]; plus 263 (an invented argument changes
scheduling — check arity before reaching for a fence the permuter cannot help with).
Cookbook 264 records the four recipes the tells agent drove to MATCH: the inline (s16)
in a call argument, the save-order/bb0 anti-dependence law (new), the opaque-bound
local assigned late (new), and 172b-1's multi-def mirror variable made concrete.
func_801F218C read as DELAY-SLOT/schedule: 83 vs 83, one mismatch, mine putting
'move a0,s0' in a jal delay slot the target leaves as nop. The 5a fence in both
placements, if/else inversion, an early-break restructuring and a 1200s permuter at
-j12 all failed. The cause was a declared parameter on a no-arg callee: passing it
makes the copy dominate the call, so the filler takes the slot. Deleting the argument
-> MATCH (83 ins).
Also records the tell that beats 195-A's 'the asm has no arity tell': an arg setup
sitting in a BRANCH delay slot, needed by only one successor, proves the call does
not use that register. And the standing waste warning — the permuter cannot change
call arity, so this class is guaranteed to defeat it.
The carve stays INSIDE the byte-gate (harvest_verify._jtbl_prep_one, the §61b-proven
order); everything new routes work to it:
* jtbl_carve: island_probe (read-only classifier: tail/covered/island-end/island-blocked/
island-pads/main-manual), --island-split (the one-line §260 insert, end-adjacent only),
and apply() now recognizes a completed island split as a no-op success instead of the
historical refusal.
* harvest_verify: _ISLAND_WALLS branch — on the §154-A refusal the gate isolates (body
still spliced), inserts the split line, re-extracts, re-carves. Snapshot-restore covers it.
* jtbl_lane.py (new): probe → draft (--draft-dir or api_agent) → gate via the exact
sweep_parallel worker call, HOLDING the campaign draw lock across gate+commit → commit
named per-binary paths. One jtbl target per gate invocation (§61c).
* build_wave_atlas: probe filter + one-jtbl-card-per-binary cap (inert unless
--levers jtbl-carve). idiom_serial: refuses the jtbl-carve lever (R43; its pre-carve ran
the refuted order — S58: 8 attempts, 0 banks).
Proven end to end with the live campaign running (banks committed separately):
commit:2661 ov_SC03_014/func_8017DCC0 tail: §8b adjacent merge + §8e pad recovery,
jr_8017AE2C.o .rodata 0x14→0x28 TIGHT, sha d84b01a2 green
commit:2663 md_SC03_076/func_801F218C covered: §260 STAGE 2 — sha 9a165e36 identical
with the table COMPILER-EMITTED; the first md_* jr bank ever
commit:2664 md_SC03_135/func_801E5358 island-end: the FULL split done BY THE GATE on a
virgin module (~1.1s, R40-checked), sha b901fda5, md.o 0x27c→0x268 + jr.o 0x14
Census (245/245 members probed, R32): 181 members / 26,445 ins reachable unattended;
main 47 parked (gate_main cadence), island-blocked 10 (stack order), island-pads 6
(needs §8e pads for modules.mk), no-jtbl 1 (atlas mislabel).
Design + failure semantics + campaign hook: docs/tool-designs/jtbl-automation-s59.md;
cookbook §260-A.
bash parses a while...done body in full before executing, so a lane-ARG change needs
a fresh shell; a python bounce re-runs the old command line (measured P31 S59: an
18-minute-old edit was invisible to the bounced process). Includes the three-way
table of what takes effect when — code / args / wave-draw defaults.
260 the 154-A leading-island split: one inserted .rodata line + jr_isolate_all --only,
the object-level sh_size control that separates 'it worked' from 'it did nothing',
the island-is-a-stack census, and the two md_*-only tool blindnesses it exposed.
261 the -O0 oracle: derive the opt level from the target (prologue tell + subseg flags),
311 $fp mentions vs 167 real -O0 prologues, and why a match is not a bank for the
116 functions stranded in -O2 subsegs.
262 a lane's yield is only a lane fact if it is size-matched — the band and size-mix
confounds that produced 'the tells lane is broken', with the bucket table.
--o0 has existed since P29 and NOTHING ever passed it: api_draft.match_one(), the
oracle every wave agent iterates against, builds a fixed argv without it. Every agent
handed an -O0 target was shown an -O2 compile of its own C and a mismatch on every
instruction — feedback that cannot converge, for a reason invisible in the diff, and
it hit even the 51 functions already sitting in -O0 objects.
match_one now decides from the target (R33), with two oracles because neither alone
is sufficient (R34): the -O0 frame-pointer prologue in the target's own bytes, OR the
subseg being an object the Makefile compiles -O0. boot/start.s is built -O0 with no
ordinary prologue; an -O0 function stranded in an -O2 subseg has the prologue but
cannot bank — and match_one now says so instead of letting an agent chase it.
The prologue scan is anchored at glabel, not the top of the file: two md_MAIN_011
.s files open with a migrated jump table / .asciz blob and a naive scan read table
words as the prologue.
tools/test_o0_detect.py: 14,400 .s scanned, 167/167 signature files covered, 0 false
positives outside -O0 objects, both rodata-first files detected.
Census correction in the design doc: the handoff's refutation of md_MAIN_003/011 is
itself wrong (16 and 21 -O0 functions, byte-verified), and the 311 '$fp mentions'
figure over-counts because $fp is $s8, allocatable at -O2. True population: 167
files, of which 116 / 14,148 ins are -O0 functions the build compiles -O2 — double
the atlas estimate, with x2/x3/x4 sibling replication.
main is the one binary splat writes to asm/nonmatchings/800/ and asm/data/ rather
than asm/<alias>/...; the hardcoded asm/<ov>/ in nine places made every main lookup
miss a directory that does not exist and report it as 'already spliced AND no stale
copy' — a true-sounding message about the wrong tree (R43).
With the path, the base and parse_config all fixed, main now reaches the real
analysis: func_8001A114 refuses because subseg 800 would host non-contiguous
.rodata carves (0x63238 = the existing LZSS table, 0x6327c = the new one), i.e.
isolate first — the same architectural refusal an overlay gets.
The island guard also moved AHEAD of jtbl_range, so an island table no longer
reports 'not found in the raw data asm — already carved / stale asm?' (true and
useless: an island table was never in the data asm). It now prints the exact line
the review prescribes: - [0x268, .rodata, md_SC03_076_jr_801F218C].
Review doc gets an S59 addendum with the measured class facts and where the two
examples now stop.
All 1,101 novel-idiom candidates read (33 waves, 4,888 verdict lines, 4,764 notes), restricted
to functions the WHOLE-BINARY BYTE-GATE banked. Append-only: 1450 insertions, 0 deletions,
prefix md5 unchanged.
§233 header+index · §234-§257 the laws · §258 the 21 addenda (append-only forbids inserting
them into their parents) · §259 the discard ledger.
Three claims were HEDGED rather than promoted (§253, §254, §213's right-rotate rule); four
were byte-verified against asm/ before being written; §217/§230/§232 were promoted from
'single observation' to CROSS-CONFIRMED by independent cards.
~1,040 discarded across 18 ledgered clusters — ~700 self-reported 'nothing the cookbook did
not already cover' (the flywheel working), ~400 Law-1c compliance reports, ~70 bare
'reordering fixed it' with no mechanism (unindexable). Two clusters deliberately left as OPEN
PROBLEMS rather than written up as false laws: forcing own_thread_p on a call-bearing join,
and suppressing a constant-offset second giv. Both are permuter jobs.
Highest-value: §234 (store lvalue signedness picks addiu vs li/ori — 12 cards each burned a
one-instruction residual on this), §236 (the nine-class declaration-layer taxonomy — ~90 cards
banked bodies that were already instruction-correct), §235 (the phantom symbol: match_one masks
relocations so a green MATCH can name a nonexistent symbol, with NO instruction diff).
The main probe (8 drafts) ran 38 minutes without a verdict. Two defects, neither about the
drafts:
1. FIXED — the compile-error shortcut matched only 'previous declaration of', but gcc printed
'previous implicit declaration of func_80017930'. So a batch whose culprit gcc had already
named fell through to bisection, which costs a full clean EXE rebuild per step. The matcher
now accepts the implicit and conflicting-types forms too. (resolve_conflicts is separately
blind to this class: an implicit decl comes from a call site with no prototype.)
2. NOT FIXED, documented — the typedef-hoist repair is not idempotent. It emitted 'hoisted 2
typedef(s)' 150 times and left a duplicated marker comment; it re-hoists, rebuilds, fails
identically and repeats, so it cannot converge. Make it idempotent and bound the bisect
before gating main again.
Also: ox_campaign pre-draws the next wave AFTER launching shards (doing it before left the
fleet at 8 agents while a card job ran), collect_drafts grants stragglers a grace period
instead of letting 2 of 220 shards idle the fleet for 34 minutes, and drafter bands are now
mostly full-range (the 400-2000 band drew 9 cards for a 2,000-worker fleet).
ox's study correctly found that the atlas o0-lane lever is MISCLASSIFIED on its probe
instance (func_801F0A9C @ md_SC03_076 has no frame pointer — verified, zero $fp under
asm/md_SC03_076/). Its proposed replacement population is fabricated: md_MAIN_003 and
md_MAIN_011 have zero $fp in either module. $fp does appear in 311 files across asm/,
which is the population an -O0 lane would really target.
Recorded rather than discarded: the misclassification finding is load-bearing (a lane
built on the lever label drafts the wrong functions) and the study named checkable files
for every claim, which is what made the refutation cheap.
Two independent reviews (ox design study + Fable validation) found parse_config does not
implement its own documented contract on md_*/main: it can DELETE the c config line and
corrupt the yaml on disk before failing. main also has no config/splat.main.yaml (it is
splat.us.exe.yaml), which is the FileNotFoundError seen on every main jtbl target.
Refuse loudly rather than corrupt quietly (R43). Lift only after parse_config is hardened
and proven on two examples.
Also: gate() stages into a per-run dir (a killed gate's 3,186 stale drafts survived into the
next gate of the same tag and poisoned every group); reloc_filter no longer treats
NOT-A-STUB as a pass (it means ALREADY BANKED — wave an staged 480 of them over source that
already byte-matches).
Designs recorded: docs/tool-designs/jtbl-island-split.md (ox) and -review.md (Fable):
14 CONFIRMED / 8 WRONG / 1 UNVERIFIABLE, verdict GO-WITH-CHANGES, and the correct fix is
SMALLER than proposed — one inserted .rodata carve line + jr_isolate_all.py --only, no _pre
piece and no ld_interleave leading mode.
R42: gate_main reverted 61 byte-proven overlay banks it could not distinguish from its own
substitution (sweep_parallel gates commit=False by design). Fixed by committing overlay banks
before the main batch, chunking main at 8 to bound bisect cost, and replacing every blind
'git checkout -- src/ config/' with commit-or-refuse in ox_campaign and idiom_serial.
R43: sweep_parallel had an explicit branch admitting main, which cannot be gated incrementally
— wave ab banked 0/105 main cards while its non-main cards banked 94/115 (82%), and the wave
read as a drafting failure. sweep_parallel now refuses main and names gate_main.py.
Also: validate_targets now prefers the card's own addr field (named symbols like SYS_OBJ_F00
were MALFORMED and discarded whole 220-card waves); ox_campaign deals model lanes by
smallest-ratio scheduling (a 73-card wave had put 73 shards on ox and 0 on deepseek);
docs/accelerators.md gains the four vacuous-check defects.
278 novel-idiom candidates extracted by tools/idiom_harvest.py from 1,004 drafting notes,
restricted to functions the WHOLE-BINARY BYTE-GATE banked (banked-ness derived from
corpus.stubs, not from match_one, whose relocation masking would admit lessons drawn from
functions we never actually reproduced).
26 new sections (§207 is the harvest header + discard ledger, §208-§232 the laws), 6 folded
as addenda to existing sections (§30, §194-B, §176-B, §165-40, §164-63, §193-A/§194-E).
~200 candidates discarded or merged — 103 of them self-reported 'nothing the cookbook did
not already cover', which is the knowledge base doing its job.
§209 CORRECTS existing guidance: §194-B's 'needs >=2 sh stores' bound is byte-wrong
(func_801A8738 has one sb and still needs the s16 declaration). Insertions only; the
existing file is byte-identical by prefix md5.
An external-model bake-off, not a banking session. Nothing was banked; that is
the next session's first task.
FINDINGS
- CARD FUEL is the biggest lever, bigger than model choice: the same 10 cards
went 4/10 -> 9/10 when seed_ref/tu_ref/decl_prior were injected. The
"60-instruction ceiling" was an artifact of withholding fuel.
- Sub-50 is near-free: 19/19 verified MATCH at $0.007/function, blind.
7,724 sub-50 open functions = 73.5% of the remaining set.
- A free model (stealth/ox-alpha) cracked a 611-ins function and a jtbl
exemplar, and distilled §206 — whose two negative results were
independently byte-confirmed before banking.
- §206 transfers WITHIN a family (40 turns -> 11) but NOT across (56 turns,
0 compiles). jtbl costs ~40 turns of learning per family, not per class.
TOOL FIXES (all negative-controlled)
- family_remap.gather_externs: took the first ALPHABETICAL extern across the
overlay's TUs, carrying two types swapped (sh/lh vs lbu/sb). Now prefers
the extracted unit's own file. Blocked a 4-member/2,444-ins family.
- atlas.member_lever: aprop_card was loaded and never read while a bare
ledger DIFF forced needs-autopsy. PURE now outranks it — rescues 32
members / 11 families / 3,810 ins.
- decl_prior._ASM_SYM: the %hi/%lo arm had never fired (0 of 1,210 over four
waves). jal 306->306 zero regressions, data 0->299.
- api_agent.py (new): --cards, --brief, --max-cost, nudge loop, 429
attribution + backoff, transport retry, non-fatal tool faults, and a
repeated-call guard.
RULES PROPOSED: R40 (exonerate the instrument before attributing a failure to
its subject — seven instances this session) and R41 (a cost/rate/yield number
ships with its denominator — I quoted $0.30 against a $6.31 bill).
Audit after "did you update tooling for those findings" found three gaps
beyond the four committed fixes:
- R21: docs/SETUP.md had no record of any of it. Adds the "four flow traps"
table — each trap, its measured cost, and where it is now caught.
- family_sweep --only had no coverage assertion, so the 3-vs-50 mis-scope
could recur silently. It now resolves member addrs to their family, always
prints the coverage line, and refuses when it resolves to zero families.
- pregate_check modelled the banking driver's typedef strip but never checked
the consequence. Adds [DROP-RISK] §203 USE-BEFORE-TYPEDEF. R39 NC: flags
the known-bad splice, 0 false positives on the post-hoist file and all 7
other wave-Z TUs (first draft read a typedef named in its own comment).
Also corrects §203 and the phase log: the text that banked was the RENAMED
variant, because gate_stage's backlog.save_draft() overwrote the original on
the failed attempt. Bytes are correct (R22 213/213); the claim "draft
byte-unchanged" was not.
Wave Z: 74/75 MATCH, 72 banked first gate, 2 recovered for zero agent tokens
(§203 address-order typedef hoist; §202 alias on a DEFINITION for the §183.3
DEF-side return wall). Sibling sweep re-scoped after --only was keyed on the
atlas gid rather than family_hseq's exemplar: 3 -> 50 banked.
Harvest: 82 gap reports -> 5 laws, 16 rejected, 30 already-covered; the
adversarial verifier killed 7 of 12 claims. §204 written (510 lines).
Tooling, all negative-controlled:
- the S46 validity gate was off the path for waves T-Z; wired into
build_wave_atlas where cards are born, with the PRE-DRAFT-ONLY constraint
documented (it condemns banked work if run later)
- wave_snapshot honors the card's `sub` (9/9 legacy byte-identical, 66/66
split-TU recovered)
- §204-E: decl_prior's %hi/%lo arm had never fired (jal 306->306 zero
regressions, data 0->299)
Refreshed rather than appended: the previous checkpoint (commit:2542) was outrun by the §201-A fix,
wave Z's redraw and wave Y's harvest, and a checkpoint the tree has moved past is worse than none
because it is believed. Written for a session that has none of this context: what banked, the seven
results that outlive the count, every tool shipped, the eight-step wave-closing sequence, the exact
invocation that fires the staged wave Z, the four known-open near-misses with their attributions,
and the transient-failure recovery that worked five times tonight.
Wave-metrics gains a single six-wave table: 405 drafted, 405 banked or recovered.
Wave Y's harvest found a defect in the card field shipped EARLIER THE SAME SESSION (§196). Overlay
functions are named by VRAM address and 134 overlays load at the same window, so build()'s bare
defs[name] counter mixed N unrelated functions -- and §196 ranked that row ABOVE the destination TU.
Measured over the tree: 3,911 of 9,861 symbols with a definition are defined in >1 binary; 1,219
disagree on ARITY, 1,204 of them in the overlay window; 818 of the disagreements were a top-two TIE
that Counter.most_common broke by sorted-file order, so the lowest-numbered defining overlay
silently won. On wave Y's five binaries, 26 of 65 overlay-window DEF rows (40%) named another
overlay's function. Byte-proven cost: applying one row's arity to func_8017E83C took it from
MATCH (114 ins) to 113 ins / 83 mismatched.
Fix: defs are keyed by BINARY; a DEF row is emitted only from the target's own binary for
overlay-window symbols, and otherwise withheld with a stated reason ('N other binaries define this
address; none is yours') rather than silently omitted. Resident/shared/main symbols are fleet-unique
and were always correct (0 of 43 wrong). build_wave_atlas passes the target binary through.
Wave Z was ALREADY STAGED with 24 poisoned rows across its 292; redrawn with the fixed tool -> 0
poisoned, 24 withheld with a reason. This is the fourth same-session self-correction (§194-E ->
§193-A, §199-A -> §189-A, §197-A -> §136-9, §201-A -> §196), and the first one caught before the
bad data reached agents.
Caught by Drew: the tool was committed and nothing called it — not the harvest prompt, not the
memory, not SETUP.md. A tool nobody invokes is the same defect this session keeps finding in the
CARDS (§193-A, §194-E, §196: the answer was already computed and nothing carried it), wearing
tooling clothes.
Now: step 0 of the harvest reader prompt runs it; docs/SETUP.md carries a row with the measured
threshold rationale (R21); the wave-closing memory names it in step 6. All three restate the same
caveat because it is the part that gets forgotten — a candidate is a STARTING POINT, and an empty
candidate list is NOT evidence of novelty, since dumb term overlap misses paraphrase entirely.
547 banked this session, fleet 96.0% instr-weighted / 91.7% distinct / 97.07% fn-count, R22 213/213
twice for wave Y. Wave Z drawn and staged at .run/wave_p31z_workflow.js + .run/wave_z_args.json
(75 cards / 6,517 ins / 8 gate groups) per Drew's instruction not to launch it. Wave Y's harvest is
the one outstanding step of the seven-step closing sequence -- three launches lost every agent to
529 Overloaded at zero tokens; gap reports and target snapshots are on disk and the checkpoint
carries the exact resume invocation.
Drew T