Wave 5 (wf_7bce6c28-9f4, 78 agents, 11.4M tok across both halves): 40 targets
-> 36 agent-MATCH, 2 refuted, 2 NEAR, 0 FAIL, 0 drafts lost -> 29 BANKED.
RESUMED THROUGH A USAGE-LIMIT OUTAGE. The first run lost 40 agents mid-wave
(18 cracks + 22 verifiers). Resuming with byte-identical args replayed the 22
completed cracks from cache and re-ran only the dead ones — a limit outage
costs the unfinished work only.
⚠ THE INTERRUPTED RESULT WAS MISLEADING AND THAT IS MY BUG: crack_wave.js
classifies anything without verdict_check.confirmed as `refuted`, so 22 drafts
whose VERIFIER DIED read as 22 refutations. A dead oracle must not be
indistinguishable from a negative verdict — split `refuted` from `unverified`.
Two prior-seeded agents did better than asked:
- func_8017F2D4: found the prior draft SURVIVED in .run/wave3/, recovered it,
then gated it IN SITU — spliced into the real destination TU, full pinned
triple, masked-diff of the function out of the whole-TU object: 0 mismatched,
279/279. That closes the §52b gap instead of trusting a standalone match. It
also byte-confirmed the x5 reach on three symbol-remapped siblings and fixed
the prior notes' TU pointer (they named a file holding only a caller).
- func_8017FFD0: VALIDATED its predecessor's claim rather than assuming it —
built a no-barrier variant and reproduced the predicted -7 ins drift exactly,
proving the __asm__ __volatile__("") cross-jump barrier is a real lever.
Bank rate across five waves: 67% -> 79% -> 69% -> 68% -> 73%.
A workflow script is plain JS; the cookbook citation I added used markdown
backticks inside a template string, which terminated it. Caught at launch
(parse error, 0 agents, 0 tokens) rather than mid-wave.
19 note-sets, 131 claims, one skeptic each, vetted against a cookbook that
already held §162/§163/§164 from this same campaign:
NEW 8 · SHARPENS 39 · COVERED 64 · UNSOUND 20
byte-probed 61 · single-instance 43 · asserted 27
COVERED+UNSOUND is 64%, up from §164's 57% — the duplicate rate RISES as the
knowledge base grows. That is the argument for harvesting after EVERY wave: a
wave launched before its predecessor's harvest lands re-derives laws already on
disk. This one was banked while wave-4 propagation was still committing.
THE PASS CORRECTED ITS OWN PREDECESSOR. §165-01 BOUNDS §163a, banked hours
earlier today. §163a says "block scope is a conflict SOLVENT" — byte-proven,
but on a DATA symbol. It does NOT reach an ARITY conflict: there the two decls
are COMPATIBLE (cc1 emits no `conflicting types` for scope to downgrade) and
the failure is call-vs-composite in convert_arguments (c-typeck.c:1623), which
a `()` declaration cannot defuse at ANY scope. The diagnostic word picks the
lever: `conflicting types` -> §163a's solvent; `too many arguments` -> cast the
call site (§17a-1/§161c) or replace the host prototype.
§165z records the 20 refuted claims. cookbook_index.py: 505 sections.
Wave 4 added 105 (27 cracks + 27/27 families propagated). Fleet 94.9% instr /
89.4% distinct / 96.44% fn-count. R22 clean-fleet run 7x this session, 213/213
every time.
Bank rate now measured four times: 67% -> 79% -> 69% -> 68%. Prior-notes
seeding 10/12 (was 7/9). func_8017C294 — the x16 family, largest item on the
board — is NEAR at 2 ins after three seeded attempts (18 -> 11 -> 2).
Also records the 4th comment-blindness defect and its blast radius (one draft
comment refused a binary's stub oracle, failed 5 later binaries, and left
drafts spliced in src/ so 17 re-gates read a poisoned tree as 0/17), and that
the wave harness now lives in tools/wave/ with its contracts written down.
The wave script encodes contracts paid for in real failures (per-agent output
dirs after wave 2 lost 21 verified drafts; sha1-last verification; prior-notes
seeding at 7/9 then 10/12; size routing). It was living only in the workflow
scratch dir, so each wave rebuilt it from memory — which is how the cookbook
citation list went stale: it still named only §162 while §163 and §164 (87 more
entries) had been banked in between.
Updated the citation list to §160-§164 and pointed agents at §164z, the 28
REFUTED claims, so they do not re-derive a disproved mechanism.
The 17 drafts that read 0/17 against the poisoned tree were never 17 codegen
failures. Re-gated on a clean tree: 9 banked, then re-gated TOGETHER per binary
(a draft that passes alone can still collide with a sibling in the same TU —
which is exactly what `morph_lerp` and `struct V8` did) — 9 of 9 held.
ov_SC03_118 func_80183D68 · ov_SC06_008 func_8017F8CC func_80183BF4
ov_SC06_011 func_8017EDA4 · ov_SC07_006 func_8017FDF8
ov_SC06_018 func_80184370 func_80184944 func_8018A974 func_8018EDB8
Wave 4 final: 40 targets -> 35 agent-MATCH -> 27 BANKED (68%).
One stray comment in one draft had masked eight real matches.
Wave 4 (wf_05895a19-121, 75 agents, 7.9M tok): 40 targets -> 35 agent-MATCH,
0 refuted, 5 NEAR, 0 FAIL, 0 drafts lost. 18 banked so far on the whole-binary
gate across 9 binaries; the other 17 are re-gating on a clean tree (see below).
PRIOR-NOTES SEEDING HELD AT SCALE: 10 of 12 seeded targets confirmed (wave 3
was 7 of 9). func_8017C294 — the x16 family, the largest single item on the
board — is now NEAR at **2 ins** (18 -> 11 -> 2 across three seeded attempts).
THE 4th COMMENT-BLINDNESS DEFECT OF THE SESSION, and the first with blast
radius. A crack agent annotated a decl in its draft:
extern void func_801842DC(s32 a0); /* TU:4023 INCLUDE_ASM (no decl) */
`corpus._INCLUDE_ASM_CAND` only skips lines that BEGIN with a comment marker,
so it read `INCLUDE_ASM (` out of the trailing PROSE, found no quoted path, and
refused the whole binary's stub oracle — correctly, by its own R32 contract.
That then failed gate_stage for every LATER binary in the run, because they all
walk the corpus: 12 binaries banked, then 5 were blocked by one comment in a
13th. Fixed the same way as the other three today: decide candidacy on
cdecl._mask'ed text, PARSE FROM THE ORIGINAL (the mask blanks string content
and would erase the asm path). Verified on 4 binaries incl. main (2,002 stubs).
AND THE DAMAGE IT LEFT: gate_stage raised out of the CorpusError BEFORE its
revert, stranding failed drafts spliced in src/. The 17 solo re-gates that
followed all read 0/17 — they were building a POISONED TREE, not judging their
own drafts. Residue reverted here; the 17 re-gate clean next.
The pattern is now sharp enough to state: any scanner that greps C source for a
token must mask comments and strings FIRST — and agent-authored drafts make it
far likelier, because their prose mentions the exact tokens our tools hunt for.
The §163z catalogue was 34 crack-agent note-sets claiming 190 distinct laws.
One independent skeptic per function, each required to read the full notes,
grep the whole cookbook, classify, and GRADE THE EVIDENCE:
NEW 20 · SHARPENS 62 · COVERED 80 · UNSOUND 28
byte-probed 114 · single-instance 51 · asserted 25
57% of what the crack agents flagged as novel was already in the cookbook or
does not survive scrutiny. That ratio is the lesson: a crack agent is the right
instrument for FINDING a lever and the wrong one for judging its novelty — it
has just spent hours in one function and has not read the other 497 sections.
Never bank a wave's flags directly.
Banked as §164-01..82, each carrying its verdict, what it sharpens, and its
evidence grade (70 byte-probed, 12 single-instance). Several skeptics CORRECTED
the mechanism the crack agent proposed while confirming its effect — e.g. the
"fold distributes the constant out of an index" claim, where the skeptic traced
the real site to expand_expr's MULT_EXPR EXPAND_SUM case (expr.c:5359-5375)
after showing pointer_int_sum's distributive law cannot fire on that tree.
§164z records the 28 REFUTED claims with the reason, so no future wave spends
tokens rediscovering them.
cookbook_index.py: 501 sections.
Drew T