Cryptanalytic Memory Lab (`cryptO1`)

A fixed-coordinate, continually learning O1 policy around resident exact search for standard full-round ChaCha20.

Author and maintainer: David Tom FOSS (DT-Foss)

cryptO1 combines three executable layers:

a native full-round ChaCha20 SAT/CDCL backend that supplies exact receipts, proof state and public verification;
O1, used as a bounded living evidence state, Bit-Vault and holographic/polyphase memory;
O1-O, used as a deterministic operator composer, policy actuator and compact failure-memory layer.

The lab is isolated from every source project. It never writes into O1, O1-O, arx-carry-leak, f8-causal-cryptanalysis or fullround-key-recovery. Published inputs cross the boundary only through pinned, SHA-256-verified, read-only adapters.

The live cockpit is STATUS.md. The strongest retained result for each attempt is in RESULTS_INDEX.md; hypotheses, exact breadcrumbs and the current resume point are under research/.

Current frontier — 2026-07-25

The current experiment is not a toy reduced-round cipher and not a progressively widened residual benchmark. It uses the standard twenty-round ChaCha20 block relation including feed-forward. All 256 key bits belong to the unknown key domain. The frozen M238 query leaves 238 candidate variables active inside that full relation; the exact M242 and M240 parent rungs were produced by the same full-round solver lineage.

The deployment-side state sees the public relation and solver-native receipts. It does not receive a revealed target key, private target round state or a candidate-indexed 2^256 table.

Proven mechanism and exact-search ladder

Layer	Retained result
Selective 256-bit retention	O1C-0020 recovers `256/256` addressed bits in all `12/12` formal cells after streams containing up to `2^20` distractors. The bounded live state is 352 bytes.
Resident living exact search	O1C-0155 proves one native PID can preserve solver state, ingest every receipt causally and change subsequent orders over hundreds of queries.
Same-query compounding	O1C-0158 converts the identical M242 query from repeated `UNKNOWN` receipts to exact `status20` in one resident PID. O1C-0161 repeats the mechanism at M240.
Causal memory-on/off discriminator	O1C-0168 replays the exact M240 identity in a fresh process: the fresh query is byte-for-byte the archived r0 `status0`, while the resident trajectory reaches exact `status20` at r106.
Exact causal attic	O1C-0170/O1C-0171 promote already-paid exact clauses to an independently enumerated union of `85,504` excluded complete keys; O1C-0178 closes the frozen inventory.
Static assumption-order causality	O1C-0232 changes only the order reaching native `solver.assume()`. In the predeclared r192..199 block, conflicts are equal and decisions fall `50,632→33,719` (`-33.4038%`); the propagation tradeoff is `+2.3360%`.
Living adaptive actuation	O1C-0237 proves the complete loop `receipt → bounded O1 state → changed order → durable publication → native solver.assume()`. Fourteen complete evidence pairs retain all 238 carriers in 237 signed levels. Complete Pair13 records the strongest completed-pair decision reduction and strongest simultaneous decision/propagation reduction among the fourteen pairs (`-12.1521%` / `-2.08469%`) at one extra conflict; cumulative propagation work is `2.224B` below static O1 control.

O1C-0237: adaptive bounded-state result

O1C-0237 is the sole live science arm. It first reproduced O1C-0213 through r199 with 8,000/8,000 exact deterministic field equalities and zero mismatches. From r200 onward, the only experimental difference is the receipt-derived order of the same 238 assumption literals. CNF, query, variables, polarities, seed and 4,096-conflict window remain fixed.

The adaptive policy has now completed fourteen complementary evidence pairs:

Pair0 closes the first full learning cycle and creates a balanced ±136,488 carrier field.
Fixed 31/32 decay plus signed phase binding compounds subsequent pair contrasts into 4, 8, 16, 32, 62, 108, 150, 218, 229, 231 and now 237 occupied signed evidence levels (234 magnitudes after fourteen completed pairs).
At the earlier exact eight-pair census (r327), all 238/238 carriers remain live. That field spans -872,580 through +668,264, reaches a smallest absolute carrier of 201, has 120 positive / 118 negative coordinates, signed sum 7, L1 59,418,325 and still uses exactly 238 state coordinates—no candidate table and no stream-length state growth.
ACK r295 durably publishes the sixth-pair full/query order at generation1994; native r296 subsequently reports updated and the exact published SHA inside solver.assume() (an actuation receipt outside the fixed r295 analysis).
Pair6 closes at r311 with exact contrast +12.6116384999%; ACK r311 publishes the resulting seventh-pair order at generation2010. Native r312..319 then proves that exact order inside solver.assume().
Pair7 closes at r327 with exact contrast -10.0894780728%; native r320..327 proves the complementary seven-pair order inside solver.assume(), and ACK r327 durably publishes the resulting eighth-pair order at generation2026.
Native r328..335 proves that eighth-pair order inside solver.assume(). Pair8's first phase is frozen at score 3580098192922/43513025; ACK r335 publishes its complementary order at generation2034 for r336.
Native r336..343 completes Pair8. Its exact +6.1474335914% contrast produces the ninth learned order, which native r344..351 then proves inside solver.assume().
Native r352..359 completes Pair9. Its exact -4.0525134701% contrast produces the tenth learned order, which native r360..367 then proves inside solver.assume().
Native r368..375 completes Pair10. Its exact -6.5172268669% contrast preserves the complete-pair propagation reduction and durably publishes the eleventh learned order for r376.
Native r376..383 proves that eleven-pair order inside solver.assume(). Conflicts and propagations both fall in the fixed block; ACK r383 publishes the complementary twelfth-pair order for r384. The ordinary r384 receipt subsequently reports selector_reload_status=updated, generation2082 and the exact published full-order SHA inside solver.assume(), completing the handoff without creating a new science boundary.
Native r384..391 proves that complementary order throughout solver.assume(). Although this second block alone costs work, the complete Pair11 retains two fewer conflicts and 207,785,211 fewer propagations with a 9,190-decision tradeoff. Its exact +6.7178556598% contrast completes the twelfth carrier update and ACK r391 publishes the twelfth learned order for r392.
Native r392..399 proves that twelfth learned order throughout solver.assume(), with generations2090..2097 and exact full/query SHA efc2ef7c…d96e / a9ee0185…1d1d. The first half of Pair12 saves one conflict but costs 15,709 decisions and 172,942,064 propagations (-0.00305%/+32.49085%/+5.16849%), route NOT_GREEN|GREEN. It is retained as an adverse half-pair, not used to prune the precommitted complementary block at r407.
Native r400..407 proves the complementary order 813c5b33…4f17 / a95b8737…f39c, generations2098..2105. This half recovers 148,490,625 propagations (-4.57744%) at +6/+6,651 conflicts/decisions. Across complete Pair12, however, the first half is not fully erased: C/D/P is +5/+22,360/+24,451,439 (+0.00763%/+23.91495%/+0.37104%). Exact contrast -4.7035999911% still performs the thirteenth bounded carrier update.
Native r408..415 proves the resulting thirteen-pair order 6e862112…e5ddf / 2e9dfc74…8778, generations2106..2113. Decisions fall by 10,862 (-18.46494%), the strongest local decision reduction among adaptive r200..415 blocks, while conflicts rise by three and propagations by 17,420,855 (+0.51237%). Both candidate and control remain NOT_GREEN; the frozen routing gate is unchanged.
Native r416..423 proves the complementary order ff31ccea…ff61 / b7389c8d…d128, generations2114..2121. That half reduces conflicts/decisions/propagations by 2/3,285/159,415,831 (-0.00610%/-5.70402%/-4.67324%), with both routes NOT_GREEN. Complete Pair13 r408..423 retains -14,147 decisions (-12.15211%) and -141,994,976 propagations (-2.08469%) at exactly one extra conflict. This is the strongest completed-pair decision reduction and strongest simultaneous decision/propagation reduction among the fourteen pairs, not an all-axis win.

The r343..r375 boundaries are authenticated retrospective exact-prefix analyses and deliberately exclude later checkpoint policy state. r383, r391 and r407/r423 landed exactly on complete checkpoints. The latest byte-identical fourteen-pair state keeps all 238/238 carriers nonzero in 237 signed levels / 234 magnitudes, with range -875,013..+778,755, minimum magnitude 913, positive/negative split 124/114, L1 61,419,400 and signed sum 2. All 238 coordinates changed relative to r407 and 38 crossed zero. This is the strongest bounded-state differentiation in the living arm, not key entropy or recovered bits.

The strict predeclared structural-work gains, the Pair5/cumulative crossover and later axis-specific field readouts are retained:

Boundary	Candidate vs static O1 control
Complete Pair3, r248..263	Equal `65,546` conflicts; decisions `92,180→85,077` (`-7.70558%`); propagations `7,377,673,172→7,357,129,434` (`-0.27846%`).
First 32-level-field block, r280..287	Equal `32,775` conflicts; decisions `44,521→40,550` (`-8.91939%`); propagations `3,845,275,892→3,761,586,697` (`-2.17642%`).
Complete Pair5, r280..295	D/P crossover with a four-conflict tradeoff: conflicts `65,548→65,552` (`+0.00610%`); decisions `103,615→94,653` (`-8.64933%`); propagations `7,206,950,757→7,148,056,632` (`-0.81719%`). This is not a third pass of H-139's equal-or-lower-conflict pair gate.
First six-pair-field block, r296..303	Equal `32,775` conflicts; propagations `3,942,913,612→3,559,012,941` (`-9.73647%`) with decisions `37,070→42,236` (`+13.93580%`). This is a large propagation reduction with an explicit decision tradeoff, not a dual-axis pass.
Complete Pair6, r296..311	Equal `65,550` conflicts; propagations `7,763,670,397→6,990,428,555` (`-9.95975%`) with decisions `81,613→92,885` (`+13.81153%`). The complementary phase reinforces the propagation/decision tradeoff and produces the seventh carrier update.
First seven-pair-field block, r312..319	Conflicts differ by `+5` (`+0.01526%`); decisions fall `40,962→37,230` (`-9.11088%`) while propagations rise `3,411,263,280→3,545,527,711` (`+3.93592%`). The seventh update flips which structural-work axis benefits.
Complete Pair7, r312..327	Conflicts differ by `+4` (`+0.00610%`); decisions fall `84,423→81,643` (`-3.29294%`) while propagations rise `7,099,277,987→7,211,361,265` (`+1.57880%`). The complementary block partly reverses r319 locally, but the complete pair retains a decision advantage and produces the eighth carrier update.
First eight-pair-field block, r328..335	Conflicts fall `32,773→32,771` (`-0.00610%`) and propagations fall `3,560,935,370→3,402,332,192` (`-4.45398%`), while decisions rise `41,587→43,418` (`+4.40282%`). This is a conflict/propagation reduction with an explicit decision tradeoff.
Complete Pair8, r328..343	Conflicts fall `65,545→65,544` (`-0.00153%`) and propagations fall `7,047,459,312→6,629,730,344` (`-5.92737%`), while decisions rise `87,016→91,623` (`+5.29443%`). The complementary phase preserves the conflict/propagation reduction across the complete pair.
First nine-pair-field block, r344..351	Conflicts are equal at `32,774`; decisions rise `45,299→52,442` (`+15.76856%`) and propagations rise `3,266,443,116→3,329,837,258` (`+1.94077%`). Candidate passes RP/RD but remains cost-contended; this block is an explicit reversal, not a gain.
Complete Pair9, r344..359	Conflicts rise by one (`+0.00153%`), decisions rise `88,564→101,100` (`+14.15474%`) and propagations fall `6,910,511,259→6,704,871,188` (`-2.97576%`). The complementary phase converts the first-block propagation cost into a complete-pair propagation reduction.
First ten-pair-field block, r360..367	Conflicts rise by five (`+0.01526%`) and decisions rise `43,028→45,364` (`+5.42902%`), while propagations fall `3,476,577,337→3,386,490,891` (`-2.59124%`).
Complete Pair10, r360..375	Conflicts rise by nine (`+0.01373%`) and decisions rise `87,881→90,595` (`+3.08827%`), while propagations fall `7,146,900,890→6,931,334,446` (`-3.01622%`). The complement preserves the propagation reduction and sharply narrows the pair's decision cost relative to Pair9.
First eleven-pair-field block, r376..383	Conflicts fall `32,776→32,773` (`-0.00915%`) and propagations fall `3,571,060,172→3,287,378,687` (`-7.94390%`), while decisions rise `46,489→52,123` (`+12.11900%`). Candidate is `NOT_GREEN` because RP/RW/K miss their unchanged gates while control is `GREEN`; the exact work shift is retained.
Complementary eleven-pair-field block, r384..391	Conflicts rise by one (`+0.00305%`), decisions rise `46,813→50,369` (`+7.59618%`) and propagations rise `3,369,469,130→3,445,365,404` (`+2.25247%`). Candidate is structurally GREEN but cost-contended; control misses RP and is `NOT_GREEN`. This adverse half is retained without threshold changes.
Complete Pair11, r376..391	Conflicts fall `65,547→65,545` (`-0.00305%`) and propagations fall `6,940,529,302→6,732,744,091` (`-2.99379%`), while decisions rise `93,302→102,492` (`+9.84974%`). The complete pair therefore preserves the first half's conflict/propagation reduction and produces the twelfth carrier update.
First twelve-pair-field block, r392..399	Conflicts fall `32,771→32,770` (`-0.00305%`), while decisions rise `48,349→64,058` (`+32.49085%`) and propagations rise `3,346,084,997→3,519,027,061` (`+5.16849%`). Candidate passes RP only and is `NOT_GREEN`; control passes all fixed guards and is `GREEN`. This is an exact adverse first-half readout, not a completed Pair12 or a pruning event.
Complementary twelve-pair-field block, r400..407	Conflicts rise `32,772→32,778` (`+0.01831%`) and decisions rise `45,149→51,800` (`+14.73122%`), while propagations fall `3,243,967,258→3,095,476,633` (`-4.57744%`). Both routes are `NOT_GREEN`; the complement reverses the first half's propagation direction without reversing its conflict/decision costs.
Complete Pair12, r392..407	Conflicts rise `65,543→65,548` (`+0.00763%`), decisions rise `93,498→115,858` (`+23.91495%`) and propagations rise `6,590,052,255→6,614,503,694` (`+0.37104%`). The complementary phase recovers most, but not all, of the first-half propagation increase; this pair is adverse on all three structural-work axes and still produces the thirteenth bounded carrier update.
First thirteen-pair-field block, r408..415	Conflicts rise `32,771→32,774` (`+0.00915%`) and propagations rise `3,400,084,400→3,417,505,255` (`+0.51237%`), while decisions fall `58,825→47,963` (`-18.46494%`). This is the strongest local decision reduction among adaptive r200..415 blocks, with explicit conflict/propagation costs; both routes remain `NOT_GREEN`.
Complementary thirteen-pair-field block, r416..423	Conflicts fall `32,774→32,772` (`-0.00610%`), decisions fall `57,591→54,306` (`-5.70402%`) and propagations fall `3,411,245,392→3,251,829,561` (`-4.67324%`). Both routes remain `NOT_GREEN`; the exact work reduction is retained without changing the frozen gate.
Complete Pair13, r408..423	Conflicts rise `65,545→65,546` (`+0.00153%`), while decisions fall `116,416→102,269` (`-12.15211%`) and propagations fall `6,811,329,792→6,669,334,816` (`-2.08469%`). This is the strongest completed-pair decision reduction and strongest simultaneous D/P reduction among the first fourteen pairs, with an explicit one-conflict tradeoff.

Across the complete adaptive prefix r200..295, candidate/control C/D/P is 393,285/510,207/43,909,758,608 versus 393,271/514,197/44,309,004,318. Candidate-minus-control is therefore +14 conflicts (+0.00356%), -3,990 decisions (-0.77597%) and -399,245,710 propagations (-0.90105%). This is the first fixed cumulative adaptive prefix in which both decisions and propagations cross below the static O1 control, with near-identical conflicts. It is a solver-work result, not yet a key-entropy or recovery result. No wall-clock speedup is claimed: cumulative candidate/control wall is 34,176,042,225/7,680,452,024 µs (+344.97%) under confounded concurrent machine load, so wall remains descriptive only.

At fixed boundary r303, the first complete block driven by that six-pair field has candidate/control C/D/P 32,775/42,236/3,559,012,941 versus 32,775/37,070/3,942,913,612: exact delta 0/+5,166/-383,900,671, or 0%/+13.93580%/-9.73647%. Across adaptive r200..303, the exact cumulative delta is now +14/+1,176/-783,146,381 (+0.00329%/+0.21333%/-1.62304%). The propagation advantage therefore strengthens while the r295 decision crossover does not persist through this next block. Both fixed observations remain valid.

Complete Pair6 at r311 reinforces the same geometry. Candidate/control C/D/P across r296..311 is 65,550/92,885/6,990,428,555 versus 65,550/81,613/7,763,670,397: exact delta 0/+11,272/-773,241,842, or 0%/+13.81153%/-9.95975%. Its frozen +/− reward scores are 3790638856092/45035675 and 3654799270247959/55954816625, yielding exact contrast 1900325998364134927687/15068034168374075002673 (+12.6116384999%, 126,116 ppm). Across adaptive r200..311, candidate-minus-control is now +14/+7,282/-1,172,487,552 (+0.00305%/+1.22220%/-2.25164%). This is sustained equal-conflict propagation reduction with a decision tradeoff, not a speedup, key-entropy or recovery result.

The first complete block under the resulting seven-pair field reaches r319. Candidate/control C/D/P is 32,779/37,230/3,545,527,711 versus 32,774/40,962/3,411,263,280: exact delta +5/-3,732/+134,264,431, or +0.01526%/-9.11088%/+3.93592%. Candidate RP/RD pass and its structural score is 2.2781957350, +6.34722% over control 2.1422240762; its canonical route remains STRUCTURAL_GREEN_COST_CONTENDED solely because load-sensitive RW/K fail, while control is NOT_GREEN on RP. Across adaptive r200..319, candidate-minus-control is +19/+3,550/-1,038,223,121 (+0.00386%/+0.55750%/-1.87121%). The new state therefore converts part of Pair6's propagation advantage into a decision advantage while cumulative propagations remain lower. At this r319 boundary Pair7 is half complete and its first-phase score is already frozen at 14806397605720064/153638651257; ACK r319 durably publishes generation2018 and the complementary full/query order for r320. The exact claim class is PROVEN_SEVEN_PAIR_FIELD_NATIVE_ACTUATION_AND_FIRST_BLOCK_DECISION_REDUCTION_CONFLICT_PROPAGATION_TRADEOFF.

Complete Pair7 is now fixed at r327. Its complementary r320..327 block has candidate/control C/D/P 32,771/44,413/3,665,833,554 versus 32,772/43,461/3,688,014,707: exact delta -1/+952/-22,181,153, or -0.00305%/+2.19047%/-0.60144%. Candidate and control are both structural GREEN; candidate structural score 2.0588437459 is +0.55802% over control 2.0474186982, while its canonical route remains STRUCTURAL_GREEN_COST_CONTENDED because load-sensitive RW/K fail. Combined with r312..319, complete Pair7 retains -2,780 decisions (-3.29294%) at +4 conflicts (+0.00610%) and +112,083,278 propagations (+1.57880%). Its frozen −/+ scores 14806397605720064/153638651257 and 1884080121922/23937883 yield exact contrast -32483142372493412114279/321950671164713610670233 (-10.0894780728%, -100,895 ppm). Fixed decay and eighth-phase binding expand the same 238-coordinate field from 108 to 150 occupied signed levels and from 62 to 109 magnitudes.

Across adaptive r200..327, candidate/control C/D/P is 524,385/684,735/58,111,548,428 versus 524,367/680,233/59,171,952,702: exact delta +18/+4,502/-1,060,404,274 (+0.00343%/+0.66183%/-1.79207%). The complete prefix therefore still carries more than one billion fewer propagations, with a small conflict and decision cost. This is exact causal search-geometry and bounded-state compounding on the frozen full-round identity—not a wall-clock speedup, Core/SAT/model, entropy reduction or key recovery claim. ACK r327 publishes generation2026 and the next full/query order SHA c6a47324b715e1f9b80bca6912f096939b2360a6071470c1d11337696b00821f / 8d38cb2d6902c3f336b7f0073a6f363150e590cb7889010dff376d1bacc362cb for r328; native observation of this new order is deliberately outside the fixed r327 artifact. The claim class advances to PROVEN_EIGHT_PAIR_150_SIGNED_LEVEL_BOUNDED_COMPOUNDING_AND_COMPLETE_PAIR_DECISION_REDUCTION_CONFLICT_PROPAGATION_TRADEOFF.

The first complete block under that eight-pair/150-level field reaches r335. Candidate/control C/D/P is 32,771/43,418/3,402,332,192 versus 32,773/41,587/3,560,935,370: exact delta -2/+1,831/-158,603,178, or -0.00610%/+4.40282%/-4.45398%. The learned field therefore reduces both conflicts and propagation work in this fixed block while spending more decisions. Candidate RD passes, but RP 1.4941218258 misses the unchanged 1.5 gate by 0.39188%; RW/K also fail, so its canonical route is NOT_GREEN. Control is GREEN, and candidate structural score 2.1050111702 is 4.94475% below control 2.2145133591. No threshold changed.

Across adaptive r200..335, candidate/control C/D/P is 557,156/728,153/61,513,880,620 versus 557,140/721,820/62,732,888,072: exact delta +16/+6,333/-1,219,007,452 (+0.00287%/+0.87737%/-1.94317%). Cumulative propagation saving therefore deepens beyond 1.2 billion while the decision overhead increases. Pair8 is only half complete; its frozen first-phase score is 3580098192922/43513025. ACK r335 durably publishes generation2034 and complementary full/query SHA fb9830b9329f712cf73a41421824e90790e893ce55268723d962aeb2ed6d8d4f / 536a4cec6826ee716b07d2380ebe897056176bdd9341bad90f164c50e776dedf for r336; native observation is outside the fixed r335 artifact. The exact claim class is PROVEN_EIGHT_PAIR_FIELD_NATIVE_ACTUATION_AND_FIRST_BLOCK_CONFLICT_PROPAGATION_REDUCTION_DECISION_TRADEOFF.

Complete Pair8 is prospectively fixed through r343. Its complementary block r336..343 changes candidate-minus-control C/D/P by +1/+2,776/-259,125,790 (+0.00305%/+6.11063%/-7.43221%). Across the complete r328..343 pair, candidate/control is 65,544/91,623/6,629,730,344 versus 65,545/87,016/7,047,459,312: exact delta -1/+4,607/-417,728,968 (-0.00153%/+5.29443%/-5.92737%). Exact +/− scores 3580098192922/43513025 and 3370713062309/46335032 yield contrast 19214042584104968779/312553886080261118229 (+6.1474335914%, 61,474 ppm). ACK r343 publishes generation2042 and the ninth learned full/query order 1307fb39…05a8 / 04a44cb2…87f9.

The first block under that order, r344..351, has equal 32,774 conflicts but adds 7,143 decisions (+15.76856%) and 63,394,142 propagations (+1.94077%). Candidate passes structural RP/RD and is STRUCTURAL_GREEN_COST_CONTENDED; control narrowly misses RP and is NOT_GREEN. This is a fixed adverse work block, retained without changing the metric. ACK r351 publishes the complementary generation2050 order 3104d197…42305 / e04733cd…ce0d9.

Complete Pair9 reaches r359. The complementary r352..359 block changes C/D/P by +1/+5,393/-269,034,213 (+0.00305%/+12.46504%/-7.38280%). Across complete Pair9 r344..359, candidate/control is 65,547/101,100/6,704,871,188 versus 65,546/88,564/6,910,511,259: exact delta +1/+12,536/-205,640,071 (+0.00153%/+14.15474%/-2.97576%). Exact +/− scores 1842913855225/26612132 and 20396690852/271591 yield contrast -42280612462203489/1043318246171029439 (-4.0525134701%, -40,525 ppm). ACK r359 publishes generation2058 and the tenth learned full/query order 37e25d44…a3c0 / b68d8066…b4b4.

Native r360..367 proves that tenth learned order in solver.assume(). Candidate/control C/D/P is 32,779/45,364/3,386,490,891 versus 32,774/43,028/3,476,577,337: exact delta +5/+2,336/-90,086,446 (+0.01526%/+5.42902%/-2.59124%). Candidate RP 1.4871077478 narrowly misses the unchanged gate and is NOT_GREEN; control is GREEN. Across the entire adaptive prefix r200..367, candidate/control is 688,255/922,822/74,832,640,851 versus 688,232/898,841/76,606,500,610: exact delta +23/+23,981/-1,773,859,759 (+0.00334%/+2.66799%/-2.31555%). The cumulative propagation saving is the largest yet, with explicit conflict and decision costs. The current claim class is PROVEN_TEN_PAIR_LIVING_ACTUATION_AND_1_773B_CUMULATIVE_PROPAGATION_REDUCTION_DECISION_TRADEOFF.

Complete Pair10 is now fixed through r375. Its complementary r368..375 block uses full/query SHA cdc94823…69fd7c / 6f7ddebd…82416 natively and changes C/D/P by +4/+378/-125,479,998 (+0.01221%/+0.84275%/-3.41877%). Candidate passes RP/RD and is STRUCTURAL_GREEN_COST_CONTENDED; control is GREEN. Across the complete r360..375 pair, candidate/control is 65,553/90,595/6,931,334,446 versus 65,544/87,881/7,146,900,890: exact delta +9/+2,714/-215,566,444 (+0.01373%/+3.08827%/-3.01622%). Exact +/− scores 2433728167055833/33339810374 and 1865097823679/22423552 yield contrast -928627979323503645/14248820829663058973 (-6.5172268669%, -65,172 ppm). ACK r375 publishes generation2074 and the eleventh learned full/query order bf06036d…32d2f / e9811b36…a8294 for r376.

Across adaptive r200..375, candidate/control C/D/P is 721,029/968,053/78,377,484,406 versus 721,002/943,694/80,276,824,163: exact delta +27/+24,359/-1,899,339,757 (+0.00374%/+2.58124%/-2.36599%). The cumulative propagation saving sets a new maximum and the decision-overhead percentage retreats from r367. The r375 claim class is PROVEN_ELEVEN_PAIR_LIVING_ACTUATION_AND_1_899B_CUMULATIVE_PROPAGATION_REDUCTION_DECISION_TRADEOFF.

The first block under that eleven-pair field is now fixed through r383. Native r376..383 proves full/query SHA bf06036d…32d2f / e9811b36…a8294 throughout. Candidate/control C/D/P is 32,773/52,123/3,287,378,687 versus 32,776/46,489/3,571,060,172: exact delta -3/+5,634/-283,681,485 (-0.00915%/+12.11900%/-7.94390%). Candidate RD passes, while RP 1.4396526831 misses the unchanged gate and the canonical route is NOT_GREEN; control is GREEN. Across adaptive r200..383, candidate/control C/D/P is 753,802/1,020,176/81,664,863,093 versus 753,778/990,183/83,847,884,335: exact delta +24/+29,993/-2,183,021,242 (+0.00318%/+3.02904%/-2.60355%).

Because r383 was captured at the exact checkpoint rather than reconstructed later, the complete 238-coordinate eleven-pair state is independently recomputable: 218 signed levels / 206 magnitudes, all nonzero, with canonical carrier-array SHA 946b9bff8ccc46f6ac94875f4934dabd6764337126084589041e37181fbcfa33. ACK r383 publishes generation2082 and the complementary r384 full/query order bb8bd769…e7ed5 / 4a70c720…e100. The fixed artifact correctly stops before native r384; the ordinary later r384 receipt confirms updated, generation2082 and exact full-order SHA bb8bd769…e7ed5 inside solver.assume(). This is continuation of the proven actuation chain, not a separate result capsule. The r383 claim class is PROVEN_ELEVEN_PAIR_218_SIGNED_LEVEL_BOUNDED_COMPOUNDING_AND_2_183B_CUMULATIVE_PROPAGATION_REDUCTION_DECISION_TRADEOFF.

The complementary r384..391 block and complete twelfth adaptive pair are now fixed. Native r384..391 proves full/query SHA bb8bd769…e7ed5 / 4a70c720…e100 throughout. That complementary block alone changes C/D/P by +1/+3,556/+75,896,274 (+0.00305%/+7.59618%/+2.25247%), route STRUCTURAL_GREEN_COST_CONTENDED|NOT_GREEN. Across complete Pair11 r376..391, candidate/control C/D/P is 65,545/102,492/6,732,744,091 versus 65,547/93,302/6,940,529,302: exact delta -2/+9,190/-207,785,211 (-0.00305%/+9.84974%/-2.99379%). Exact minus/plus scores 3449583481102/53953393 and 3710053589533057/50721690038 yield contrast 25201275278477036150525/375138683455798385474277 (+6.7178556598%, 67,179 ppm).

Across adaptive r200..391, candidate/control C/D/P is 786,574/1,070,545/85,110,228,497 versus 786,549/1,036,996/87,217,353,465: exact delta +25/+33,549/-2,107,124,968 (+0.00318%/+3.23521%/-2.41595%). The cumulative propagation advantage remains exact but retreats from the r383 maximum; no monotone-improvement claim is made.

The exact r391 checkpoint closes twelve pairs in the same 238 coordinates: 238/238 carriers remain nonzero in 229 signed levels / 223 magnitudes, with canonical carrier-array SHA 1d1c4f56bad6818e04a0b7be61f8aa45054a05912cf60e020c57abee827586ed. ACK r391 publishes generation2090 and r392 full/query order efc2ef7c…d96e / a9ee0185…1d1d. Native observation is outside the fixed artifact; the ordinary later r392 receipt confirms updated, generation2090 and exact full-order SHA efc2ef7c…d96e inside solver.assume(), without opening a new science boundary. The strongest current claim class is PROVEN_TWELVE_PAIR_229_SIGNED_LEVEL_BOUNDED_COMPOUNDING_AND_COMPLETE_PAIR_CONFLICT_PROPAGATION_REDUCTION_DECISION_TRADEOFF.

The first block under that twelve-pair field is now fixed at r399. Its candidate/control C/D/P totals are 32,770/64,058/3,519,027,061 versus 32,771/48,349/3,346,084,997, exact delta -1/+15,709/+172,942,064 (-0.003051%/+32.490848%/+5.168490%). Candidate RP/RD/RW/K is 1.518767888598/1.957684928996/2.081603411691/0.828359983617; only RP passes, so the frozen route is NOT_GREEN, while the static control passes all four gates and is GREEN. Across adaptive r200..399, candidate/control C/D/P is 819,344/1,134,603/88,629,255,558 versus 819,320/1,085,345/90,563,438,462: delta +24/+49,258/-1,934,182,904 (+0.002929%/+4.538465%/-2.135722%). Thus the cumulative propagation advantage remains exact but shrinks again; neither it nor the first-half conflict reduction is promoted to monotone improvement.

The r399 boundary contains 400 status0 receipts, zero Core/model/terminal and 208 complete durable ACK→native proofs. Pair12 is deliberately incomplete: its +score is 3639153166439/65917280, its −score is null and the twelve-pair carrier state remains unchanged. ACK r399 durably publishes generation2098 and r400 full/query order 813c5b33…4f17 / a95b8737…f39c; native observation is pending inside the fixed artifact. The ordinary later r400 status0 receipt now confirms selector_reload_status=updated, generation2098 and exact expected/observed/assumption full-order SHA 813c5b33…4f17 inside solver.assume(); this is mechanism continuity, not a new science boundary. No second policy snapshot is needed because r391 already freezes the byte-identical twelve-pair field. The precommitted complementary block through r407—not this half alone—determines complete Pair12. The strongest complete-pair claim therefore remains the r391 class above.

Complete Pair12 is now fixed at r407. Native r400..407 proves full/query SHA 813c5b33…4f17 / a95b8737…f39c throughout. Candidate/control C/D/P in that complementary block is 32,778/51,800/3,095,476,633 versus 32,772/45,149/3,243,967,258: exact delta +6/+6,651/-148,490,625 (+0.018308%/+14.731223%/-4.577439%), route NOT_GREEN|NOT_GREEN. Thus the complement does reverse the first half's propagation direction, validating the precommitted pair-level decision unit, but it does not recover the complete pair.

Across r392..407, candidate/control C/D/P is 65,548/115,858/6,614,503,694 versus 65,543/93,498/6,590,052,255: exact delta +5/+22,360/+24,451,439 (+0.007629%/+23.914950%/+0.371036%). Complete Pair12 is therefore adverse on all three structural-work axes. Its exact score contrast -40065398272181895959509/851802839268558741040171 (-4.7035999911%, -47,036 ppm) nevertheless executes the frozen thirteenth carrier update; a work null does not erase the bounded-state transition.

Across adaptive r200..407, candidate/control C/D/P is 852,122/1,186,403/91,724,732,191 versus 852,092/1,130,494/93,807,405,720: exact delta +30/+55,909/-2,082,673,529 (+0.003521%/+4.945537%/-2.220159%). The cumulative propagation advantage rebounds from r399 but remains below the r383 maximum, so no monotonicity or new cumulative maximum is claimed.

The exact r407 checkpoint freezes thirteen complete pairs in the same 238 coordinates. All carriers remain nonzero in 231 signed levels / 226 magnitudes, positive/negative 116/122, range -836,592..+656,915, minimum magnitude 218, L1 54,354,123, signed sum 5, canonical carrier SHA 09ec4c13006ccd286d23dd2d953a6673d14008e193bef562b3cd2b9365468663. All 238 coordinates change relative to r391 and 12 cross zero; the transition is a field-wide living update, not a renamed copy of the prior state. This supersedes r391 only for bounded-state differentiation; r391 remains the strongest local complete-pair work gain and r383 retains the cumulative propagation maximum. The exact r407 class is PROVEN_THIRTEEN_PAIR_231_SIGNED_LEVEL_BOUNDED_COMPOUNDING_AND_2_083B_CUMULATIVE_PROPAGATION_REDUCTION_DECISION_TRADEOFF. ACK r407 publishes generation2106 and r408 full/query order 6e862112…e5ddf / 2e9dfc74…8778; native observation is pending inside the fixed artifact. The ordinary later r408 status0 receipt now confirms selector_reload_status=updated, generation2106 and identical expected/observed/assumption full-order SHA 6e862112…e5ddf inside solver.assume(); this is actuation continuity, not a new science boundary.

The first full block driven by the thirteen-pair field is now fixed at r415. Exact r0..415 contains 416 status0 receipts, zero Core/model/terminal and 224 complete durable ACK→native links. Native r408..415 authenticates full/query SHA 6e862112…e5ddf / 2e9dfc74…8778 at generations2106..2113. Candidate/control C/D/P is 32,774/47,963/3,417,505,255 versus 32,771/58,825/3,400,084,400: exact delta +3/-10,862/+17,420,855 (+0.0091544353%/-18.4649383765%/+0.5123653695%), route NOT_GREEN|NOT_GREEN. The decision reduction is the strongest local one among adaptive r200..415 blocks. It is an axis-specific result with conflict and propagation costs, not a reason to change the frozen routing gate.

Across adaptive r200..415, candidate/control C/D/P is 884,896/1,234,366/95,142,237,446 versus 884,863/1,189,319/97,207,490,120: exact delta +33/+45,047/-2,065,252,674 (+0.003729390877%/+3.787629727600%/-2.124581831555%). Pair13 remains deliberately incomplete after its negative first half: minus score 3610095713863275/48711134008, plus score and contrast null. Therefore the 238 carriers remain byte-for-byte the r407 field with canonical SHA 09ec4c13…8663, and no redundant state snapshot is published. The retained hierarchy is unchanged: r383 has the maximum cumulative propagation advantage, r391 the strongest positive complete-pair result, r407 the strongest bounded-state differentiation, and r415 the strongest single-block decision reduction.

ACK r415 publishes generation2114 and pending r416 full/query order ff31ccea…ff61 / b7389c8d…d128; native r416 is pending inside the fixed artifact. Outside that boundary, the ordinary later r416 status0 receipt confirms the exact full/query order in native solver.assume() from the selector, and ACK r416 republishes it at generation2115 for r417. This is continuity only, not a new capsule or science boundary. There is no recovery, terminal, speedup or cross-identity generalization claim; the next fixed boundary is r423.

Complete Pair13—the fourteenth adaptive pair—is now fixed at r423. Exact non-retrospective r0..423 contains 424 status0 receipts, zero Core/model/terminal and 232 complete durable ACK→native links. Native r416..423 authenticates full/query SHA ff31ccea…ff61 / b7389c8d…d128 at generations2114..2121. Complementary-block candidate/control C/D/P is 32,772/54,306/3,251,829,561 versus 32,774/57,591/3,411,245,392: exact delta -2/-3,285/-159,415,831 (-0.0061023982%/-5.7040162525%/-4.6732443047%), route NOT_GREEN|NOT_GREEN. Walltime remains descriptive only under uncontrolled concurrent load and is excluded from the claim.

Across complete Pair13 r408..423, candidate/control C/D/P is 65,546/102,269/6,669,334,816 versus 65,545/116,416/6,811,329,792: exact delta +1/-14,147/-141,994,976 (+0.001525669387%/-12.152109675646%/-2.084688017409%). This is the strongest completed-pair decision reduction and strongest simultaneous decision/propagation reduction among the fourteen completed pairs. The one extra conflict remains explicit, so this is not an all-axis win. Frozen minus/plus scores 3610095713863275/48711134008 and 3375592803987/60668872 yield exact contrast -1137322527930592141373/7988528920907726878452 (-14.236945740460%, -142,369 ppm).

Across adaptive r200..423, candidate/control C/D/P is 917,668/1,288,672/98,394,067,007 versus 917,637/1,246,910/100,618,735,512: exact delta +31/+41,762/-2,224,668,505 (+0.003378242159%/+3.349239319598%/-2.210988335005%). This is a new cumulative propagation maximum, 41,647,263 better than r383, with explicit conflict/decision tradeoffs.

The exact r423 checkpoint freezes fourteen complete pairs in the same 238 coordinates. All carriers remain nonzero in 237 signed levels / 234 magnitudes; positive/negative is 124/114, range -875,013..+778,755, minimum magnitude 913, L1 61,419,400, signed sum 2, canonical carrier SHA 343d428fedb88e25c96f7b6b3f73fe5210f2cface3f8f8e157ae8c7deee0c915. All 238 coordinates changed relative to r407 and 38 crossed zero, making r423 the strongest bounded-state differentiation. The retained hierarchy is now r415 for the strongest single-block decision reduction, r391 for the retained positive-contrast complete-pair designation, and r423 for cumulative propagation, complete-pair decision and bounded-state records. The exact r423 class is PROVEN_FOURTEEN_PAIR_237_SIGNED_LEVEL_BOUNDED_COMPOUNDING_AND_COMPLETE_PAIR13_DECISION_PROPAGATION_REDUCTION_ONE_CONFLICT_TRADEOFF.

ACK r423 publishes generation2122 and pending r424 full/query order c17a92bf…e74a / 1b7c1e68…e521. Outside the fixed artifact, ordinary r424 status0 confirms that exact selector order inside native solver.assume(); ACK r424 republishes it at generation2123. This is continuity only, not a new capsule or science boundary. There is no recovery, terminal, speedup or cross-identity generalization claim. The next fixed boundary is r431; r439 then completes Pair14.

The fixed r295 artifact is O1C0237_R295_SIXTH_COMPLETE_ADAPTIVE_PAIR_ANALYSIS (95,988 bytes, mode 0444, SHA-256 abb4eff6b8937cf120102cb32d7a1da600114d073694b28b1fcfc0ed4fcef9e2). The complete 238-coordinate state and its independently recomputable level census are frozen in O1C0237_R295_SIX_PAIR_CARRIER_CENSUS (5,006 bytes, mode 0444, SHA-256 ddcb54522df1bfaa7ec8eaacf5f95d3d89c9974766c73dcbfd388080764546c5). The ordinary post-boundary native link is frozen separately as O1C0237_R296_SIX_PAIR_NATIVE_ACTUATION_WITNESS (5,854 bytes, mode 0444, SHA-256 c888d9c256915608e18f078b655df812816e750ed3832ebd737416069c17b4b4); it is explicitly mechanism evidence, not a new science boundary. The latest fixed r303 six-pair-field analysis is 102,481 bytes, mode 0444, SHA-256 dd03fbd4f09b3350afcb2af349f74c7885546e62fa84f7203e2c34a783607cb2. The latest fixed r311 complete-Pair6 analysis is 109,147 bytes, mode 0444, SHA-256 c2f4296ebb621beb665307cc74bfe76b989f6aa6e4258fdde76073cd2434bf49. Its complete seventh-pair field is independently frozen in the r311 carrier census (4,874 bytes, mode 0444, SHA-256 6d78a01e5216ce4c9760248ecff2aea49e30248ca2a445adec63722033be1c54). The latest fixed r319 seven-pair-field analysis is 115,641 bytes, mode 0444, SHA-256 c80c2615381b0d40483cd3cf756753a1342d9f6d1037d52e4ece6678efdedf61. The latest fixed r327 complete-Pair7 analysis is 122,327 bytes, mode 0444, SHA-256 2bb09d86afdd1d8a10580c3507d4d32669c0b7deeb1c79c1c5f9cef2e8152481. Its exact eight-pair policy state is frozen independently in the r327 policy-state snapshot (7,457 bytes, mode 0444, SHA-256 ba554fe920822841824beb77ca2caf60fce28143668447e6da5fdffb1116eda7); the canonical compact carrier array hashes to 260b255455fde3d4bc835a7d3b344c505a79016f381d497599b1d4e05774a2c2. The fixed r335 eight-pair-field analysis is 128,710 bytes, mode 0444, SHA-256 510471d7f9b9ca364438bbdb8964220450980396bf2f6665225fa2ab10c31668. The authenticated retrospective exact-prefix r343 complete-Pair8 analysis is 133,268 bytes, mode 0444, SHA-256 64e87b961cf7508ab05c6b9b1cabdc90454db83731539bbf31a3c02c8315c665. The corresponding r351 nine-pair-field analysis is 139,765 bytes, mode 0444, SHA-256 3eca224e982a3553c2d6f3fea69ea8c8b8c07b96d0dbf45bb36be13a1dab550a; the r359 complete-Pair9 analysis is 146,274 bytes, mode 0444, SHA-256 abcc6fc3d8c639a02d823bfa102ab02e4a30a52de7bb5ad060ac4e96a128fd2a; and the r367 ten-pair-field analysis is 152,743 bytes, mode 0444, SHA-256 9c336b2c770bc2b5a94507cd80cd82cd524f937b6e9cb4202047875ed6add102. The r375 complete-Pair10 analysis is 159,217 bytes, mode 0444, SHA-256 9568ef798405171ee00369e1bc26dc3e4699f7c32b1189803899f054020eb8c1. Those r343..r375 artifacts exclude later candidate/ACK rows before invoking the sealed analyzer, supply zero post-boundary bytes to it and include no later checkpoint policy state. The exact-checkpoint r383 eleven-pair-field analysis is 167,563 bytes, mode 0444, SHA-256 9645af1b5f5bfd1a822157f24742df52906204da144c974b0b467cdd4e5eb9f9. Its byte-identical eleven-pair policy-state snapshot is 7,419 bytes, mode 0444, SHA-256 f3fefc8bfd5ad47b84d0fe5e1878c6788e6ad3455cbf51b50b5d99e10493e70d. The exact-checkpoint r391 complete-Pair11 analysis is 174,140 bytes, mode 0444, SHA-256 940175ef76cc65ee581a5337bdb26ab7c1f42a6c95b9d3eef3b11ab6284963d9. Its byte-identical twelve-pair policy-state snapshot is 7,476 bytes, mode 0444, SHA-256 a03ad3de675ac6ab6b5e8fd7b245418ec5ed63e4bd200ab1b96b96f5602cb1c3. The exact-checkpoint r399 first twelve-pair-field analysis is 180,575 bytes, mode 0444, SHA-256 2a1fffaff5f972e3eb1049a0229a2242b48afbef67bbdd7eddc36842d8fe691d. It copies exactly the 400-receipt checkpoint boundary, supplies zero post-boundary bytes to the sealed analyzer and records zero process signals, runtime writes or target/reveal reads. The exact-checkpoint r407 complete-Pair12 analysis is 187,102 bytes, mode 0444, SHA-256 51a749ebbe1acad46b2228ff9964451e38d8b7d696bcc379bfbb1a226f8c9df8. Its byte-identical thirteen-pair policy-state snapshot is 7,478 bytes, mode 0444, SHA-256 818c0835a5d4189be0f43f6a806925c63440c6c1736199588afbbd2b6c95e01c; the canonical compact carrier array hashes to 09ec4c13006ccd286d23dd2d953a6673d14008e193bef562b3cd2b9365468663. The exact-checkpoint r415 first thirteen-pair-field analysis is 193,491 bytes, mode 0444, SHA-256 8d7281a24a1843486272baac6f2404992bfd628dfdf72cd195e776fb49486970. It copies the 416-receipt checkpoint boundary, supplies zero post-boundary bytes to the sealed analyzer and records zero process signals, runtime writes or target/reveal reads. The exact-checkpoint r423 complete-Pair13 analysis is 199,994 bytes, mode 0444, SHA-256 bbf04032a98dd120da7c1967081d2027e12e5481052f1aeae0f3716fbe73bcfb. Its byte-identical fourteen-pair policy-state snapshot is 7,460 bytes, mode 0444, SHA-256 d2af1d8dc8f0b79a05fa784267d81ab7fc27e9bdb8bf844c26b24277222caa2b; the canonical compact carrier array hashes to 343d428fedb88e25c96f7b6b3f73fe5210f2cface3f8f8e157ae8c7deee0c915. Both copy exactly the 424-receipt checkpoint boundary and record zero post-boundary analyzer bytes, process signals, runtime writes or target/reveal reads. The uninterrupted process continues toward fixed r431, complete Pair14 at r439, the frozen 19-pair horizon at r503, exploit r504..511 and its natural terminal.

Architecture

flowchart LR
    A["Public full-round ChaCha20 relation"] --> B["Native resident CDCL solver"]
    B --> C["Exact receipt stream"]
    C --> D["Bounded O1 evidence state"]
    D --> E["Holographic carrier binding"]
    E --> F["O1-O deterministic order composer"]
    F --> G["Durable selector publication"]
    G --> H["solver.assume()"]
    H --> B
    C --> I["Surprise-selected exact causal attic"]
    I --> D

The loop lives in one resident solver process. O1 retains a fixed 238-coordinate carrier field, a bounded current block, two pair rewards and an eight-entry failure-hash ring. O1-O deterministically composes the carrier field with the public base order and fails closed on missing, duplicate or incorrect variables. The exact attic grows only through separately accounted, surprise-selected clauses; it is not hidden inside the O(1)-in-stream state claim. Native CDCL clauses/proof state are likewise separate and may grow; only the named O1 policy/evidence state is fixed in stream length.

The implementation surface is public:

Experimental contract

r0..191 is Natural Order.
r192..199 uses one frozen public O1 order.
r200..503 contains nineteen predeclared complementary adaptive probe pairs.
r504..511 is the predeclared exploit block.
Every query uses the same full-round CNF, M238 query, 238 variables, candidate-derived polarities, seed and conflict budget.
The controller must fsync the order, then metadata, then ACK before issuing the next command.
Every native receipt records the exact selector generation and assumption order SHA that reached solver.assume().
Scientific interpretation occurs only at fixed 8-query block, 16-query pair, Core/SAT/model or natural-terminal boundaries.
A status0 receipt means only UNKNOWN in that resident state and budget.

The live arm uses one nice-10 native CPU thread, no GPU/MPS/Metal and remains well below its 1 GiB RSS contract. The archived O1C-0236 prefix is an implementation-lifecycle defect and is never resumed; O1C-0237 changes only the process host while preserving the science byte-for-byte.

Active Moonshot: O1-256 Living Inverse

The terminal target remains exact recovery of a fresh unknown 256-bit key from standard full-round public inputs and output. The intended path is:

public target → solver-native contrast stream → bounded O1 state → addressable 256-bit posterior/beam → exact public ChaCha20 verification.

Progress is measured at several non-substitutable layers: exact retention, causal memory-on/off effects, exact domain exclusions, per-bit entropy, beam rank, time-to-hit and finally a publicly verified key. A useful intermediate result is not discarded because it is smaller than full recovery, and it is not renamed into recovery either.

The complete attacker contract is in O1-256 Living Inverse. The W52 mechanisms are mapped in the transfer report, and the continual fast/slow learning design is in O1 Online Möbius Controller.

Public disclosure and claim boundary

This repository is the defensive publication of the cryptO1 research line under Apache-2.0. It publishes source, frozen configurations, exact positive and negative results, immutable boundary artifacts and the claim ledger. Git history is part of the prior-art record.

The repository currently establishes:

exact learned 256/256 long-stream retention in bounded live state;
executable resident O1 exact search and same-query proof-state compounding;
an exact causal M240 memory-on/off discriminator;
exact Full-Round domain exclusions and an 85,504-key causal attic;
native assumption-order causality under a matched full-round relation;
adaptive fourteen-pair holographic evidence compounding through 237 occupied signed levels / 234 magnitudes in 238 fixed coordinates, with the preceding living order authenticated through native r423 and 232 complete durable ACK→native proofs;
prospectively fixed blocks with simultaneous decision and propagation reduction, the first cumulative adaptive prefix with both axes below static O1 control, and a later complete equal-conflict pair that exchanges +13.81153% decisions for -9.95975% propagations; the next complete pair retains -3.29294% decisions at a +1.57880% propagation tradeoff, and its resulting field produces a fixed -4.45398% propagation / +4.40282% decision shift with two fewer conflicts; complete Pair8 retains a -5.92737% propagation and one-conflict reduction, and cumulative complete Pair10 retains -3.01622% propagations at +3.08827% decisions; complete Pair11 retains -2.99379% propagations and two fewer conflicts at a +9.84974% decision tradeoff; complete Pair12 is adverse at +0.00763%/+23.91495%/+0.37104% C/D/P even though its complementary half reverses propagation direction, while cumulative r200..407 retains -2,082,673,529 propagations (-2.22016%) with explicit conflict/decision tradeoffs; complete Pair13 then records -12.15211% decisions and -2.08469% propagations at one extra conflict, while cumulative r200..423 reaches a new maximum of 2,224,668,505 fewer propagations (-2.21099%) and the bounded state reaches 237 signed levels.

It does not yet claim an attacker-valid recovery of a fresh unknown ChaCha20 key or a practical break of ChaCha20. That boundary is exact and does not downgrade the mechanisms and search gains already proven.

See the defensive publication, claims and limits, release notes, reproduction guide, citation metadata, current status and results index.

What the first benchmark proves

The initial benchmark deliberately separates three questions that are easy to confound:

Closed-gate storage qualification. Once relevance has already been decided, can a fixed state retain 256 binary bindings through a long haystack?
Evidence amplification. If each public observation already contains a weak, independent bit signal, does a streaming accumulator amplify it? Does the same apparent per-observation accuracy fail when errors are correlated?
Information-flow safety. Can a typed operator chain produce a frozen target-blind order while structurally rejecting any path that uses a revealed model or target secret?

It does not claim that a full-round cipher exposes such a signal. O1C-0018 is the first direct observability test and does not pass its raw-signal gate.

The full-context attention arm is an explicitly invalid O(T) attack but an exact harness ceiling. The direct 256-register vault is an intentionally honest bounded baseline. It is constant in stream length, but it is a position-indexed register bank and therefore does not demonstrate holographic compression. The holographic arm receives the same number of scalar cells (128 complex channels = 256 real scalars), while the undersized CountSketch arm is a capacity control. Precision and serialized byte size are still reported separately before any efficiency claim; equal cell count is not equal information budget.

The current haystack transition is an explicit ideal no-op for bounded arms. This isolates storage capacity and holographic crosstalk; it does not claim to have trained O1's selective input gate. Learned unified-token routing is the next memory stage, not a hidden property of this smoke test.

Run

The package pins NumPy 2.2.6 because exact reproduction of the historical Direct12 floating-point reader is part of the evidence contract; the remaining harness uses the Python standard library.

python3 -m venv .venv
.venv/bin/pip install -r requirements-dev.txt
.venv/bin/pip install -e .

# Current bounded-memory and adaptive-assumption path
PYTHONPATH=src .venv/bin/python -m pytest -q \
  tests/test_selective_mqar.py \
  tests/test_selective_mqar_run.py \
  scratch/o1c0210_assumption_feed_crossover/test_isolation.py \
  scratch/o1c0236_adaptive_assumption_actuation/test_adaptive_order_policy.py \
  scratch/o1c0236_adaptive_assumption_actuation/test_adaptive_assumption_controller.py \
  scratch/o1c0236_adaptive_assumption_actuation/test_analyze_o1c0236.py \
  scratch/o1c0237_launchd_adaptive_assumption_actuation/test_analyze_o1c0237.py

# Verify the strongest dual-axis boundary and latest fixed boundary
artifact=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R295_SIXTH_COMPLETE_ADAPTIVE_PAIR_ANALYSIS_20260725T073930+0200.json
census=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R295_SIX_PAIR_CARRIER_CENSUS_20260725T074951+0200.json
witness=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R296_SIX_PAIR_NATIVE_ACTUATION_WITNESS_20260725T075720+0200.json
field6=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R303_FIRST_SIX_PAIR_FIELD_BLOCK_ANALYSIS_20260725T082019+0200.json
pair7=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R311_SEVENTH_COMPLETE_ADAPTIVE_PAIR_ANALYSIS_20260725T085942+0200.json
census7=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R311_SEVEN_PAIR_CARRIER_CENSUS_20260725T090110+0200.json
field7=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R319_FIRST_SEVEN_PAIR_FIELD_BLOCK_ANALYSIS_20260725T094113+0200.json
pair8=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R327_EIGHTH_COMPLETE_ADAPTIVE_PAIR_ANALYSIS_20260725T102954+0200.json
state8=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R327_EIGHT_PAIR_POLICY_STATE_SNAPSHOT_20260725T102954+0200.json
field8=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R335_FIRST_EIGHT_PAIR_FIELD_BLOCK_ANALYSIS_20260725T111559+0200.json
pair9=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R343_NINTH_COMPLETE_ADAPTIVE_PAIR_ANALYSIS_20260725T151925+0200.json
field9=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R351_FIRST_NINE_PAIR_FIELD_BLOCK_ANALYSIS_20260725T151925+0200.json
pair10=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R359_TENTH_COMPLETE_ADAPTIVE_PAIR_ANALYSIS_20260725T151925+0200.json
field10=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R367_FIRST_TEN_PAIR_FIELD_BLOCK_ANALYSIS_20260725T151926+0200.json
pair11=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R375_ELEVENTH_COMPLETE_ADAPTIVE_PAIR_ANALYSIS_20260725T153444+0200.json
field11=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R383_FIRST_ELEVEN_PAIR_FIELD_BLOCK_ANALYSIS_20260725T161057+0200.json
state11=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R383_ELEVEN_PAIR_POLICY_STATE_SNAPSHOT_20260725T161057+0200.json
pair12=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R391_TWELFTH_COMPLETE_ADAPTIVE_PAIR_ANALYSIS_20260725T165809+0200.json
state12=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R391_TWELVE_PAIR_POLICY_STATE_SNAPSHOT_20260725T165809+0200.json
field12=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R399_FIRST_TWELVE_PAIR_FIELD_BLOCK_ANALYSIS_20260725T174629+0200.json
pair13=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R407_THIRTEENTH_COMPLETE_ADAPTIVE_PAIR_ANALYSIS_20260725T182836+0200.json
state13=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R407_THIRTEEN_PAIR_POLICY_STATE_SNAPSHOT_20260725T182836+0200.json
field13=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R415_FIRST_THIRTEEN_PAIR_FIELD_BLOCK_ANALYSIS_20260725T191217+0200.json
pair14=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R423_FOURTEENTH_COMPLETE_ADAPTIVE_PAIR_ANALYSIS_20260725T195315+0200.json
state14=scratch/o1c0237_launchd_adaptive_assumption_actuation/O1C0237_R423_FOURTEEN_PAIR_POLICY_STATE_SNAPSHOT_20260725T195315+0200.json
test "$(shasum -a 256 "$artifact" | cut -d ' ' -f 1)" = \
  abb4eff6b8937cf120102cb32d7a1da600114d073694b28b1fcfc0ed4fcef9e2
test "$(shasum -a 256 "$census" | cut -d ' ' -f 1)" = \
  ddcb54522df1bfaa7ec8eaacf5f95d3d89c9974766c73dcbfd388080764546c5
test "$(shasum -a 256 "$witness" | cut -d ' ' -f 1)" = \
  c888d9c256915608e18f078b655df812816e750ed3832ebd737416069c17b4b4
test "$(shasum -a 256 "$field6" | cut -d ' ' -f 1)" = \
  dd03fbd4f09b3350afcb2af349f74c7885546e62fa84f7203e2c34a783607cb2
test "$(shasum -a 256 "$pair7" | cut -d ' ' -f 1)" = \
  c2f4296ebb621beb665307cc74bfe76b989f6aa6e4258fdde76073cd2434bf49
test "$(shasum -a 256 "$census7" | cut -d ' ' -f 1)" = \
  6d78a01e5216ce4c9760248ecff2aea49e30248ca2a445adec63722033be1c54
test "$(shasum -a 256 "$field7" | cut -d ' ' -f 1)" = \
  c80c2615381b0d40483cd3cf756753a1342d9f6d1037d52e4ece6678efdedf61
test "$(shasum -a 256 "$pair8" | cut -d ' ' -f 1)" = \
  2bb09d86afdd1d8a10580c3507d4d32669c0b7deeb1c79c1c5f9cef2e8152481
test "$(shasum -a 256 "$state8" | cut -d ' ' -f 1)" = \
  ba554fe920822841824beb77ca2caf60fce28143668447e6da5fdffb1116eda7
test "$(shasum -a 256 "$field8" | cut -d ' ' -f 1)" = \
  510471d7f9b9ca364438bbdb8964220450980396bf2f6665225fa2ab10c31668
test "$(shasum -a 256 "$pair9" | cut -d ' ' -f 1)" = \
  64e87b961cf7508ab05c6b9b1cabdc90454db83731539bbf31a3c02c8315c665
test "$(shasum -a 256 "$field9" | cut -d ' ' -f 1)" = \
  3eca224e982a3553c2d6f3fea69ea8c8b8c07b96d0dbf45bb36be13a1dab550a
test "$(shasum -a 256 "$pair10" | cut -d ' ' -f 1)" = \
  abcc6fc3d8c639a02d823bfa102ab02e4a30a52de7bb5ad060ac4e96a128fd2a
test "$(shasum -a 256 "$field10" | cut -d ' ' -f 1)" = \
  9c336b2c770bc2b5a94507cd80cd82cd524f937b6e9cb4202047875ed6add102
test "$(shasum -a 256 "$pair11" | cut -d ' ' -f 1)" = \
  9568ef798405171ee00369e1bc26dc3e4699f7c32b1189803899f054020eb8c1
test "$(shasum -a 256 "$field11" | cut -d ' ' -f 1)" = \
  9645af1b5f5bfd1a822157f24742df52906204da144c974b0b467cdd4e5eb9f9
test "$(shasum -a 256 "$state11" | cut -d ' ' -f 1)" = \
  f3fefc8bfd5ad47b84d0fe5e1878c6788e6ad3455cbf51b50b5d99e10493e70d
test "$(shasum -a 256 "$pair12" | cut -d ' ' -f 1)" = \
  940175ef76cc65ee581a5337bdb26ab7c1f42a6c95b9d3eef3b11ab6284963d9
test "$(shasum -a 256 "$state12" | cut -d ' ' -f 1)" = \
  a03ad3de675ac6ab6b5e8fd7b245418ec5ed63e4bd200ab1b96b96f5602cb1c3
test "$(shasum -a 256 "$field12" | cut -d ' ' -f 1)" = \
  2a1fffaff5f972e3eb1049a0229a2242b48afbef67bbdd7eddc36842d8fe691d
test "$(shasum -a 256 "$pair13" | cut -d ' ' -f 1)" = \
  51a749ebbe1acad46b2228ff9964451e38d8b7d696bcc379bfbb1a226f8c9df8
test "$(shasum -a 256 "$state13" | cut -d ' ' -f 1)" = \
  818c0835a5d4189be0f43f6a806925c63440c6c1736199588afbbd2b6c95e01c
test "$(shasum -a 256 "$field13" | cut -d ' ' -f 1)" = \
  8d7281a24a1843486272baac6f2404992bfd628dfdf72cd195e776fb49486970
test "$(shasum -a 256 "$pair14" | cut -d ' ' -f 1)" = \
  bbf04032a98dd120da7c1967081d2027e12e5481052f1aeae0f3716fbe73bcfb
test "$(shasum -a 256 "$state14" | cut -d ' ' -f 1)" = \
  d2af1d8dc8f0b79a05fa784267d81ab7fc27e9bdb8bf844c26b24277222caa2b
jq -e '
  .candidate_endpoint.receipt_count == 296 and
  .candidate_endpoint.status_counts["0"] == 296 and
  .candidate_policy_state.completed_pairs == 6 and
  .candidate_policy_state.carrier_summary.count == 238 and
  .candidate_policy_state.carrier_summary.nonzero == 238 and
  .aggregate.cumulative_delta_candidate_minus_control.decisions == -3990 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -399245710 and
  .assumption_actuation.pending_next_actuation.full_order_sha256 ==
    "a53d242b8a51140ec1a47b9f6223eab30f84607eb83ad19c6df8cc78ace5508e" and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$artifact"
jq -e '
  (.carriers | length) == 238 and
  ([.carriers[] | select(. != 0)] | length) == 238 and
  ([.carriers[]] | unique | length) == 62 and
  ([.carriers[] | if . < 0 then -. else . end] | unique | length) == 32 and
  ([.carriers[] | if . < 0 then -. else . end] | add) == 56945052
' "$census"
jq -e '
  .classification == "MECHANISM_LINK_ONLY_NOT_SCIENCE_BOUNDARY" and
  .ack_r295.selector_generation == 1994 and
  .receipt_r296.expected_selector_generation == 1994 and
  .receipt_r296.observed_selector_generation == 1994 and
  .receipt_r296.expected_selector_order_sha256 ==
    .receipt_r296.observed_selector_order_sha256 and
  .receipt_r296.observed_selector_order_sha256 ==
    .receipt_r296.assumption_order_sha256 and
  .receipt_r296.selector_reload_status == "updated"
' "$witness"
jq -e '
  .candidate_endpoint.receipt_count == 304 and
  .candidate_endpoint.status_counts["0"] == 304 and
  .candidate_endpoint.core_count == 0 and
  .candidate_endpoint.model_count == 0 and
  .matched_blocks[-1].sequence_start == 296 and
  .matched_blocks[-1].sequence_end == 303 and
  .matched_blocks[-1].candidate.totals.conflicts == 32775 and
  .matched_blocks[-1].control.totals.conflicts == 32775 and
  (.matched_blocks[-1].candidate.totals.decisions -
    .matched_blocks[-1].control.totals.decisions) == 5166 and
  (.matched_blocks[-1].candidate.totals.propagations -
    .matched_blocks[-1].control.totals.propagations) == -383900671 and
  .aggregate.cumulative_delta_candidate_minus_control.decisions == 1176 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -783146381 and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$field6"
jq -e '
  .candidate_endpoint.receipt_count == 312 and
  .candidate_endpoint.status_counts["0"] == 312 and
  .candidate_policy_state.completed_pairs == 7 and
  .candidate_policy_state.carrier_summary.nonzero == 238 and
  .aggregate.cumulative_delta_candidate_minus_control.decisions == 7282 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -1172487552 and
  .assumption_actuation.pending_next_actuation.full_order_sha256 ==
    "1745ca4a22a89da8303c62297b5173a03273c3a84d2015e0004741d61ed16e2a" and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$pair7"
jq -e '
  (.carriers | length) == 238 and
  ([.carriers[] | select(. != 0)] | length) == 238 and
  ([.carriers[]] | unique | length) == 108 and
  ([.carriers[] | if . < 0 then -. else . end] | unique | length) == 62 and
  ([.carriers[] | if . < 0 then -. else . end] | add) == 58442222
' "$census7"
jq -e '
  .classification == "MATCHED_ADAPTIVE_PREFIX" and
  .candidate_endpoint.receipt_count == 320 and
  .candidate_endpoint.status_counts["0"] == 320 and
  .candidate_endpoint.status_counts["10"] == 0 and
  .candidate_endpoint.status_counts["20"] == 0 and
  .candidate_endpoint.core_count == 0 and
  .candidate_endpoint.model_count == 0 and
  .candidate_endpoint.terminal == false and
  .matched_blocks[-1].sequence_start == 312 and
  .matched_blocks[-1].sequence_end == 319 and
  (.matched_blocks[-1].candidate.totals.conflicts -
    .matched_blocks[-1].control.totals.conflicts) == 5 and
  (.matched_blocks[-1].candidate.totals.decisions -
    .matched_blocks[-1].control.totals.decisions) == -3732 and
  (.matched_blocks[-1].candidate.totals.propagations -
    .matched_blocks[-1].control.totals.propagations) == 134264431 and
  .aggregate.cumulative_delta_candidate_minus_control.conflicts == 19 and
  .aggregate.cumulative_delta_candidate_minus_control.decisions == 3550 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -1038223121 and
  .assumption_actuation.completed_actuation_proof_count == 128 and
  ([.assumption_actuation.selector_native_links[] |
    select(.sequence >= 312 and .sequence <= 319) |
    select(.actuation_proof_complete == true and
      .full_order_sha256 == "1745ca4a22a89da8303c62297b5173a03273c3a84d2015e0004741d61ed16e2a" and
      .query_order_sha256 == "a1488c7fa110d747030ebe91f602a0a70424df16db1988e2634438981c60521b" and
      .query_order_count == 238)] | length) == 8 and
  .candidate_policy_state.completed_pairs == 7 and
  .candidate_policy_state.current_pair_index == 7 and
  .candidate_policy_state.current_pair_minus_score ==
    "14806397605720064/153638651257" and
  .candidate_policy_state.current_pair_plus_score == null and
  .assumption_actuation.pending_next_actuation.sequence == 320 and
  .assumption_actuation.pending_next_actuation.selector_generation == 2018 and
  .assumption_actuation.pending_next_actuation.full_order_sha256 ==
    "f84ef074d1fddb15252676614c2fa869a3113c7ff8783b3edb2b0037c51d012b" and
  .boundary.next_fixed_boundary_sequence_end == 327 and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$field7"
jq -e '
  .classification == "MATCHED_ADAPTIVE_PREFIX" and
  .candidate_endpoint.receipt_count == 328 and
  .candidate_endpoint.status_counts["0"] == 328 and
  .candidate_endpoint.status_counts["10"] == 0 and
  .candidate_endpoint.status_counts["20"] == 0 and
  .candidate_endpoint.core_count == 0 and
  .candidate_endpoint.model_count == 0 and
  .candidate_endpoint.terminal == false and
  .matched_blocks[-1].sequence_start == 320 and
  .matched_blocks[-1].sequence_end == 327 and
  (.matched_blocks[-1].candidate.totals.conflicts -
    .matched_blocks[-1].control.totals.conflicts) == -1 and
  (.matched_blocks[-1].candidate.totals.decisions -
    .matched_blocks[-1].control.totals.decisions) == 952 and
  (.matched_blocks[-1].candidate.totals.propagations -
    .matched_blocks[-1].control.totals.propagations) == -22181153 and
  .aggregate.cumulative_delta_candidate_minus_control.conflicts == 18 and
  .aggregate.cumulative_delta_candidate_minus_control.decisions == 4502 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -1060404274 and
  .assumption_actuation.completed_actuation_proof_count == 136 and
  ([.assumption_actuation.selector_native_links[] |
    select(.sequence >= 320 and .sequence <= 327) |
    select(.actuation_proof_complete == true and
      .full_order_sha256 == "f84ef074d1fddb15252676614c2fa869a3113c7ff8783b3edb2b0037c51d012b" and
      .query_order_sha256 == "8b0cf7955efc9d3f9989a098f8d0622071aa7742fc6af34b8e692d594582ff67" and
      .query_order_count == 238)] | length) == 8 and
  .candidate_policy_state.completed_pairs == 8 and
  .candidate_policy_state.carrier_summary.count == 238 and
  .candidate_policy_state.carrier_summary.nonzero == 238 and
  .candidate_policy_state.carrier_summary.minimum == -872580 and
  .candidate_policy_state.carrier_summary.maximum == 668264 and
  .candidate_policy_state.carrier_summary.l1 == 59418325 and
  .candidate_policy_state.last_policy_event.contrast ==
    "-32483142372493412114279/321950671164713610670233" and
  .candidate_policy_state.last_policy_event.contrast_ppm == -100895 and
  .assumption_actuation.pending_next_actuation.sequence == 328 and
  .assumption_actuation.pending_next_actuation.selector_generation == 2026 and
  .assumption_actuation.pending_next_actuation.full_order_sha256 ==
    "c6a47324b715e1f9b80bca6912f096939b2360a6071470c1d11337696b00821f" and
  .assumption_actuation.pending_next_actuation.query_order_sha256 ==
    "8d38cb2d6902c3f336b7f0073a6f363150e590cb7889010dff376d1bacc362cb" and
  .boundary.next_fixed_boundary_sequence_end == 335 and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$pair8"
jq -e '
  .attempt_count == 328 and
  .last_attempt.sequence == 327 and
  .policy_state.completed_pairs == 8 and
  .policy_state.next_sequence == 328 and
  (.policy_state.carriers | length) == 238 and
  ([.policy_state.carriers[] | select(. != 0)] | length) == 238 and
  ([.policy_state.carriers[]] | unique | length) == 150 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | unique | length) == 109 and
  ([.policy_state.carriers[] | select(. > 0)] | length) == 120 and
  ([.policy_state.carriers[] | select(. < 0)] | length) == 118 and
  ([.policy_state.carriers[]] | add) == 7 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | min) == 201 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | add) == 59418325 and
  .selector_generation == 2026 and
  .selector_order_sha256 ==
    "c6a47324b715e1f9b80bca6912f096939b2360a6071470c1d11337696b00821f" and
  .target_or_reveal_consumed == false and
  .terminal == null
' "$state8"
test "$(jq -c -j '.policy_state.carriers' "$state8" | shasum -a 256 | cut -d ' ' -f 1)" = \
  260b255455fde3d4bc835a7d3b344c505a79016f381d497599b1d4e05774a2c2
jq -e '
  .classification == "MATCHED_ADAPTIVE_PREFIX" and
  .candidate_endpoint.receipt_count == 336 and
  .candidate_endpoint.status_counts["0"] == 336 and
  .candidate_endpoint.status_counts["10"] == 0 and
  .candidate_endpoint.status_counts["20"] == 0 and
  .candidate_endpoint.core_count == 0 and
  .candidate_endpoint.model_count == 0 and
  .candidate_endpoint.terminal == false and
  .matched_blocks[-1].sequence_start == 328 and
  .matched_blocks[-1].sequence_end == 335 and
  (.matched_blocks[-1].candidate.totals.conflicts -
    .matched_blocks[-1].control.totals.conflicts) == -2 and
  (.matched_blocks[-1].candidate.totals.decisions -
    .matched_blocks[-1].control.totals.decisions) == 1831 and
  (.matched_blocks[-1].candidate.totals.propagations -
    .matched_blocks[-1].control.totals.propagations) == -158603178 and
  .aggregate.cumulative_delta_candidate_minus_control.conflicts == 16 and
  .aggregate.cumulative_delta_candidate_minus_control.decisions == 6333 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -1219007452 and
  .assumption_actuation.completed_actuation_proof_count == 144 and
  ([.assumption_actuation.selector_native_links[] |
    select(.sequence >= 328 and .sequence <= 335) |
    select(.actuation_proof_complete == true and
      .full_order_sha256 == "c6a47324b715e1f9b80bca6912f096939b2360a6071470c1d11337696b00821f" and
      .query_order_sha256 == "8d38cb2d6902c3f336b7f0073a6f363150e590cb7889010dff376d1bacc362cb" and
      .query_order_count == 238)] | length) == 8 and
  .candidate_policy_state.completed_pairs == 8 and
  .candidate_policy_state.current_pair_index == 8 and
  .candidate_policy_state.current_pair_plus_score ==
    "3580098192922/43513025" and
  .candidate_policy_state.current_pair_minus_score == null and
  .assumption_actuation.pending_next_actuation.sequence == 336 and
  .assumption_actuation.pending_next_actuation.selector_generation == 2034 and
  .assumption_actuation.pending_next_actuation.full_order_sha256 ==
    "fb9830b9329f712cf73a41421824e90790e893ce55268723d962aeb2ed6d8d4f" and
  .assumption_actuation.pending_next_actuation.query_order_sha256 ==
    "536a4cec6826ee716b07d2380ebe897056176bdd9341bad90f164c50e776dedf" and
  .boundary.next_fixed_boundary_sequence_end == 343 and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$field8"
jq -e '
  .classification == "MATCHED_ADAPTIVE_PREFIX" and
  .candidate_endpoint.receipt_count == 344 and
  .candidate_endpoint.status_counts["0"] == 344 and
  .candidate_endpoint.core_count == 0 and
  .candidate_endpoint.model_count == 0 and
  .candidate_endpoint.terminal == false and
  .matched_blocks[-1].sequence_start == 336 and
  .matched_blocks[-1].sequence_end == 343 and
  (([.matched_blocks[-2:][] | .candidate.totals.conflicts] | add) -
    ([.matched_blocks[-2:][] | .control.totals.conflicts] | add)) == -1 and
  (([.matched_blocks[-2:][] | .candidate.totals.decisions] | add) -
    ([.matched_blocks[-2:][] | .control.totals.decisions] | add)) == 4607 and
  (([.matched_blocks[-2:][] | .candidate.totals.propagations] | add) -
    ([.matched_blocks[-2:][] | .control.totals.propagations] | add)) == -417728968 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -1478133242 and
  .assumption_actuation.completed_actuation_proof_count == 152 and
  .assumption_actuation.pending_next_actuation.sequence == 344 and
  .assumption_actuation.pending_next_actuation.selector_generation == 2042 and
  .assumption_actuation.pending_next_actuation.full_order_sha256 ==
    "1307fb3999e2110b41416bb9d1c89e9156f611f906e8902496ec309189fc05a8" and
  .bounded_snapshot.retrospective_prefix == true and
  .bounded_snapshot.post_boundary_candidate_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.post_boundary_ack_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.later_checkpoint_policy_state_included_in_result == false and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$pair9"
jq -e '
  .candidate_endpoint.receipt_count == 352 and
  .matched_blocks[-1].sequence_start == 344 and
  .matched_blocks[-1].sequence_end == 351 and
  (.matched_blocks[-1].candidate.totals.conflicts -
    .matched_blocks[-1].control.totals.conflicts) == 0 and
  (.matched_blocks[-1].candidate.totals.decisions -
    .matched_blocks[-1].control.totals.decisions) == 7143 and
  (.matched_blocks[-1].candidate.totals.propagations -
    .matched_blocks[-1].control.totals.propagations) == 63394142 and
  .assumption_actuation.completed_actuation_proof_count == 160 and
  .bounded_snapshot.retrospective_prefix == true and
  .bounded_snapshot.post_boundary_candidate_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.post_boundary_ack_bytes_supplied_to_sealed_analyzer == 0
' "$field9"
jq -e '
  .candidate_endpoint.receipt_count == 360 and
  .matched_blocks[-1].sequence_start == 352 and
  .matched_blocks[-1].sequence_end == 359 and
  (([.matched_blocks[-2:][] | .candidate.totals.conflicts] | add) -
    ([.matched_blocks[-2:][] | .control.totals.conflicts] | add)) == 1 and
  (([.matched_blocks[-2:][] | .candidate.totals.decisions] | add) -
    ([.matched_blocks[-2:][] | .control.totals.decisions] | add)) == 12536 and
  (([.matched_blocks[-2:][] | .candidate.totals.propagations] | add) -
    ([.matched_blocks[-2:][] | .control.totals.propagations] | add)) == -205640071 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -1683773313 and
  .assumption_actuation.completed_actuation_proof_count == 168 and
  .bounded_snapshot.retrospective_prefix == true and
  .bounded_snapshot.post_boundary_candidate_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.post_boundary_ack_bytes_supplied_to_sealed_analyzer == 0
' "$pair10"
jq -e '
  .candidate_endpoint.receipt_count == 368 and
  .candidate_endpoint.status_counts["0"] == 368 and
  .candidate_endpoint.status_counts["10"] == 0 and
  .candidate_endpoint.status_counts["20"] == 0 and
  .candidate_endpoint.core_count == 0 and
  .candidate_endpoint.model_count == 0 and
  .candidate_endpoint.terminal == false and
  .matched_blocks[-1].sequence_start == 360 and
  .matched_blocks[-1].sequence_end == 367 and
  (.matched_blocks[-1].candidate.totals.conflicts -
    .matched_blocks[-1].control.totals.conflicts) == 5 and
  (.matched_blocks[-1].candidate.totals.decisions -
    .matched_blocks[-1].control.totals.decisions) == 2336 and
  (.matched_blocks[-1].candidate.totals.propagations -
    .matched_blocks[-1].control.totals.propagations) == -90086446 and
  .aggregate.cumulative_delta_candidate_minus_control.conflicts == 23 and
  .aggregate.cumulative_delta_candidate_minus_control.decisions == 23981 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -1773859759 and
  .assumption_actuation.completed_actuation_proof_count == 176 and
  .assumption_actuation.pending_next_actuation.sequence == 368 and
  .assumption_actuation.pending_next_actuation.selector_generation == 2066 and
  .assumption_actuation.pending_next_actuation.full_order_sha256 ==
    "cdc94823427e7b89a3725d6eefb9e99c7a375c6f6b68411c4038d0a4ce69fd7c" and
  .boundary.next_fixed_boundary_sequence_end == 375 and
  .bounded_snapshot.retrospective_prefix == true and
  .bounded_snapshot.post_boundary_candidate_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.post_boundary_ack_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.later_checkpoint_policy_state_included_in_result == false and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$field10"
jq -e '
  .candidate_endpoint.receipt_count == 376 and
  .candidate_endpoint.status_counts["0"] == 376 and
  .candidate_endpoint.status_counts["10"] == 0 and
  .candidate_endpoint.status_counts["20"] == 0 and
  .candidate_endpoint.core_count == 0 and
  .candidate_endpoint.model_count == 0 and
  .candidate_endpoint.terminal == false and
  .matched_blocks[-1].sequence_start == 368 and
  .matched_blocks[-1].sequence_end == 375 and
  (.matched_blocks[-1].candidate.totals.conflicts -
    .matched_blocks[-1].control.totals.conflicts) == 4 and
  (.matched_blocks[-1].candidate.totals.decisions -
    .matched_blocks[-1].control.totals.decisions) == 378 and
  (.matched_blocks[-1].candidate.totals.propagations -
    .matched_blocks[-1].control.totals.propagations) == -125479998 and
  (([.matched_blocks[-2:][] | .candidate.totals.conflicts] | add) -
    ([.matched_blocks[-2:][] | .control.totals.conflicts] | add)) == 9 and
  (([.matched_blocks[-2:][] | .candidate.totals.decisions] | add) -
    ([.matched_blocks[-2:][] | .control.totals.decisions] | add)) == 2714 and
  (([.matched_blocks[-2:][] | .candidate.totals.propagations] | add) -
    ([.matched_blocks[-2:][] | .control.totals.propagations] | add)) == -215566444 and
  .aggregate.cumulative_delta_candidate_minus_control.conflicts == 27 and
  .aggregate.cumulative_delta_candidate_minus_control.decisions == 24359 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -1899339757 and
  .assumption_actuation.completed_actuation_proof_count == 184 and
  .assumption_actuation.pending_next_actuation.sequence == 376 and
  .assumption_actuation.pending_next_actuation.selector_generation == 2074 and
  .assumption_actuation.pending_next_actuation.full_order_sha256 ==
    "bf06036db7c825630d72d6bb6f296d859e19faf20bfd6cf55851c0cc97332d2f" and
  .assumption_actuation.pending_next_actuation.query_order_sha256 ==
    "e9811b36af4e2f1218ff7b3c630040862cd676380d228db6180006d923fa8294" and
  .boundary.next_fixed_boundary_sequence_end == 383 and
  .bounded_snapshot.retrospective_prefix == true and
  .bounded_snapshot.candidate_receipts_excluded_after_selected_boundary == 1 and
  .bounded_snapshot.ack_rows_excluded_after_selected_boundary == 1 and
  .bounded_snapshot.post_boundary_candidate_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.post_boundary_ack_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.later_checkpoint_policy_state_included_in_result == false and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$pair11"
jq -e '
  .candidate_endpoint.receipt_count == 384 and
  .candidate_endpoint.status_counts["0"] == 384 and
  .candidate_endpoint.status_counts["10"] == 0 and
  .candidate_endpoint.status_counts["20"] == 0 and
  .candidate_endpoint.core_count == 0 and
  .candidate_endpoint.model_count == 0 and
  .candidate_endpoint.terminal == false and
  .matched_blocks[-1].sequence_start == 376 and
  .matched_blocks[-1].sequence_end == 383 and
  (.matched_blocks[-1].candidate.totals.conflicts -
    .matched_blocks[-1].control.totals.conflicts) == -3 and
  (.matched_blocks[-1].candidate.totals.decisions -
    .matched_blocks[-1].control.totals.decisions) == 5634 and
  (.matched_blocks[-1].candidate.totals.propagations -
    .matched_blocks[-1].control.totals.propagations) == -283681485 and
  .aggregate.cumulative_delta_candidate_minus_control.conflicts == 24 and
  .aggregate.cumulative_delta_candidate_minus_control.decisions == 29993 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -2183021242 and
  .assumption_actuation.completed_actuation_proof_count == 192 and
  .assumption_actuation.pending_next_actuation.sequence == 384 and
  .assumption_actuation.pending_next_actuation.selector_generation == 2082 and
  .assumption_actuation.pending_next_actuation.full_order_sha256 ==
    "bb8bd769207acee3bb40977a592e3f32dafe6081e91b7a12f2f9caf9ae9e7ed5" and
  .assumption_actuation.pending_next_actuation.query_order_sha256 ==
    "4a70c720f4ec6de3ec49da4617b2368a81512061c10f4bdefa2b6de71441e100" and
  .boundary.next_fixed_boundary_sequence_end == 391 and
  .bounded_snapshot.retrospective_prefix == false and
  .bounded_snapshot.checkpoint_attempt_count == 384 and
  .bounded_snapshot.later_checkpoint_policy_state_included_in_result == true and
  .bounded_snapshot.post_boundary_candidate_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.post_boundary_ack_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$field11"
jq -e '
  .attempt_count == 384 and
  .last_attempt.sequence == 383 and
  .policy_state.completed_pairs == 11 and
  .policy_state.next_sequence == 384 and
  (.policy_state.carriers | length) == 238 and
  ([.policy_state.carriers[] | select(. != 0)] | length) == 238 and
  ([.policy_state.carriers[]] | unique | length) == 218 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | unique | length) == 206 and
  ([.policy_state.carriers[] | select(. > 0)] | length) == 114 and
  ([.policy_state.carriers[] | select(. < 0)] | length) == 124 and
  ([.policy_state.carriers[]] | min) == -825084 and
  ([.policy_state.carriers[]] | max) == 586217 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | min) == 1042 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | add) == 57227466 and
  ([.policy_state.carriers[]] | add) == 12 and
  .selector_generation == 2082 and
  .selector_order_sha256 ==
    "bb8bd769207acee3bb40977a592e3f32dafe6081e91b7a12f2f9caf9ae9e7ed5" and
  .target_or_reveal_consumed == false and
  .terminal == null
' "$state11"
test "$(jq -c -j '.policy_state.carriers' "$state11" | shasum -a 256 | cut -d ' ' -f 1)" = \
  946b9bff8ccc46f6ac94875f4934dabd6764337126084589041e37181fbcfa33
jq -e '
  .candidate_endpoint.receipt_count == 392 and
  .candidate_endpoint.status_counts["0"] == 392 and
  .candidate_endpoint.status_counts["10"] == 0 and
  .candidate_endpoint.status_counts["20"] == 0 and
  .candidate_endpoint.core_count == 0 and
  .candidate_endpoint.model_count == 0 and
  .candidate_endpoint.terminal == false and
  .matched_blocks[-1].sequence_start == 384 and
  .matched_blocks[-1].sequence_end == 391 and
  (.matched_blocks[-1].candidate.totals.conflicts -
    .matched_blocks[-1].control.totals.conflicts) == 1 and
  (.matched_blocks[-1].candidate.totals.decisions -
    .matched_blocks[-1].control.totals.decisions) == 3556 and
  (.matched_blocks[-1].candidate.totals.propagations -
    .matched_blocks[-1].control.totals.propagations) == 75896274 and
  (([.matched_blocks[-2:][] | .candidate.totals.conflicts] | add) -
    ([.matched_blocks[-2:][] | .control.totals.conflicts] | add)) == -2 and
  (([.matched_blocks[-2:][] | .candidate.totals.decisions] | add) -
    ([.matched_blocks[-2:][] | .control.totals.decisions] | add)) == 9190 and
  (([.matched_blocks[-2:][] | .candidate.totals.propagations] | add) -
    ([.matched_blocks[-2:][] | .control.totals.propagations] | add)) == -207785211 and
  .aggregate.cumulative_delta_candidate_minus_control.conflicts == 25 and
  .aggregate.cumulative_delta_candidate_minus_control.decisions == 33549 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -2107124968 and
  .candidate_policy_state.completed_pairs == 12 and
  .candidate_policy_state.last_policy_event.pair_complete == true and
  .candidate_policy_state.last_policy_event.contrast_ppm == 67179 and
  .assumption_actuation.completed_actuation_proof_count == 200 and
  .assumption_actuation.pending_next_actuation.sequence == 392 and
  .assumption_actuation.pending_next_actuation.selector_generation == 2090 and
  .assumption_actuation.pending_next_actuation.full_order_sha256 ==
    "efc2ef7c47232b88592aa6536bd449fb08319d45a1896aa474de77120eafd96e" and
  .assumption_actuation.pending_next_actuation.query_order_sha256 ==
    "a9ee0185b79ec834e9f9b2f07c4ca6230ce7280baef558478fbf41e4042f1d1d" and
  .boundary.next_fixed_boundary_sequence_end == 399 and
  .bounded_snapshot.retrospective_prefix == false and
  .bounded_snapshot.checkpoint_attempt_count == 392 and
  .bounded_snapshot.later_checkpoint_policy_state_included_in_result == true and
  .bounded_snapshot.post_boundary_candidate_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.post_boundary_ack_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$pair12"
jq -e '
  .attempt_count == 392 and
  .last_attempt.sequence == 391 and
  .policy_state.completed_pairs == 12 and
  .policy_state.next_sequence == 392 and
  (.policy_state.carriers | length) == 238 and
  ([.policy_state.carriers[] | select(. != 0)] | length) == 238 and
  ([.policy_state.carriers[]] | unique | length) == 229 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | unique | length) == 223 and
  ([.policy_state.carriers[] | select(. > 0)] | length) == 118 and
  ([.policy_state.carriers[] | select(. < 0)] | length) == 120 and
  ([.policy_state.carriers[]] | min) == -815026 and
  ([.policy_state.carriers[]] | max) == 629552 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | min) == 5707 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | add) == 54596843 and
  ([.policy_state.carriers[]] | add) == 7 and
  .selector_generation == 2090 and
  .selector_order_sha256 ==
    "efc2ef7c47232b88592aa6536bd449fb08319d45a1896aa474de77120eafd96e" and
  .target_or_reveal_consumed == false and
  .terminal == null
' "$state12"
test "$(jq -c -j '.policy_state.carriers' "$state12" | shasum -a 256 | cut -d ' ' -f 1)" = \
  1d1c4f56bad6818e04a0b7be61f8aa45054a05912cf60e020c57abee827586ed
jq -e '
  .candidate_endpoint.receipt_count == 400 and
  .candidate_endpoint.status_counts["0"] == 400 and
  .candidate_endpoint.status_counts["10"] == 0 and
  .candidate_endpoint.status_counts["20"] == 0 and
  .candidate_endpoint.core_count == 0 and
  .candidate_endpoint.model_count == 0 and
  .candidate_endpoint.terminal == false and
  .matched_blocks[-1].sequence_start == 392 and
  .matched_blocks[-1].sequence_end == 399 and
  .matched_blocks[-1].candidate.totals.conflicts == 32770 and
  .matched_blocks[-1].candidate.totals.decisions == 64058 and
  .matched_blocks[-1].candidate.totals.propagations == 3519027061 and
  .matched_blocks[-1].control.totals.conflicts == 32771 and
  .matched_blocks[-1].control.totals.decisions == 48349 and
  .matched_blocks[-1].control.totals.propagations == 3346084997 and
  .matched_blocks[-1].candidate.score.routing_state == "NOT_GREEN" and
  .matched_blocks[-1].control.score.routing_state == "GREEN" and
  .aggregate.cumulative_delta_candidate_minus_control.conflicts == 24 and
  .aggregate.cumulative_delta_candidate_minus_control.decisions == 49258 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -1934182904 and
  .candidate_policy_state.completed_pairs == 12 and
  .candidate_policy_state.current_pair_index == 12 and
  .candidate_policy_state.current_pair_plus_score == "3639153166439/65917280" and
  .candidate_policy_state.current_pair_minus_score == null and
  .candidate_policy_state.last_policy_event.pair_complete == false and
  .assumption_actuation.completed_actuation_proof_count == 208 and
  .assumption_actuation.selector_native_link_count == 208 and
  .assumption_actuation.selector_native_links[-1].sequence == 399 and
  .assumption_actuation.selector_native_links[-1].selector_generation == 2097 and
  .assumption_actuation.selector_native_links[-1].full_order_sha256 ==
    "efc2ef7c47232b88592aa6536bd449fb08319d45a1896aa474de77120eafd96e" and
  .assumption_actuation.selector_native_links[-1].query_order_sha256 ==
    "a9ee0185b79ec834e9f9b2f07c4ca6230ce7280baef558478fbf41e4042f1d1d" and
  .assumption_actuation.pending_next_actuation.sequence == 400 and
  .assumption_actuation.pending_next_actuation.selector_generation == 2098 and
  .assumption_actuation.pending_next_actuation.full_order_sha256 ==
    "813c5b33790640f7e10ca7c3e0cc28691c8db6c99df2924128acf90a87c84f17" and
  .assumption_actuation.pending_next_actuation.query_order_sha256 ==
    "a95b873710e2d9c567ce8f80a109dd74d7bb9e66233a79c7be4febbfb392f39c" and
  .boundary.next_fixed_boundary_sequence_end == 407 and
  .bounded_snapshot.mode == "copied-checkpoint-boundary-complete-lines" and
  .bounded_snapshot.retrospective_prefix == false and
  .bounded_snapshot.checkpoint_attempt_count == 400 and
  .bounded_snapshot.post_boundary_candidate_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.post_boundary_ack_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$field12"
jq -e '
  .candidate_endpoint.receipt_count == 408 and
  .candidate_endpoint.status_counts["0"] == 408 and
  .candidate_endpoint.status_counts["10"] == 0 and
  .candidate_endpoint.status_counts["20"] == 0 and
  .candidate_endpoint.core_count == 0 and
  .candidate_endpoint.model_count == 0 and
  .candidate_endpoint.terminal == false and
  .matched_blocks[-1].sequence_start == 400 and
  .matched_blocks[-1].sequence_end == 407 and
  (.matched_blocks[-1].candidate.totals.conflicts -
    .matched_blocks[-1].control.totals.conflicts) == 6 and
  (.matched_blocks[-1].candidate.totals.decisions -
    .matched_blocks[-1].control.totals.decisions) == 6651 and
  (.matched_blocks[-1].candidate.totals.propagations -
    .matched_blocks[-1].control.totals.propagations) == -148490625 and
  (([.matched_blocks[-2:][] | .candidate.totals.conflicts] | add) -
    ([.matched_blocks[-2:][] | .control.totals.conflicts] | add)) == 5 and
  (([.matched_blocks[-2:][] | .candidate.totals.decisions] | add) -
    ([.matched_blocks[-2:][] | .control.totals.decisions] | add)) == 22360 and
  (([.matched_blocks[-2:][] | .candidate.totals.propagations] | add) -
    ([.matched_blocks[-2:][] | .control.totals.propagations] | add)) == 24451439 and
  .aggregate.cumulative_delta_candidate_minus_control.conflicts == 30 and
  .aggregate.cumulative_delta_candidate_minus_control.decisions == 55909 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -2082673529 and
  .candidate_policy_state.completed_pairs == 13 and
  .candidate_policy_state.last_policy_event.pair_complete == true and
  .candidate_policy_state.last_policy_event.pair_index == 12 and
  .candidate_policy_state.last_policy_event.contrast_ppm == -47036 and
  .assumption_actuation.completed_actuation_proof_count == 216 and
  .assumption_actuation.selector_native_link_count == 216 and
  .assumption_actuation.pending_next_actuation.sequence == 408 and
  .assumption_actuation.pending_next_actuation.selector_generation == 2106 and
  .assumption_actuation.pending_next_actuation.full_order_sha256 ==
    "6e862112c25932cc69c8319637bb8cc17a4bb759c2f49b7d4e3936d5743e5ddf" and
  .assumption_actuation.pending_next_actuation.query_order_sha256 ==
    "2e9dfc74b6c11ce881d2684446b4aef1341bbf14bf5d76567381469332768778" and
  .boundary.next_fixed_boundary_sequence_end == 415 and
  .bounded_snapshot.mode == "copied-checkpoint-boundary-complete-lines" and
  .bounded_snapshot.retrospective_prefix == false and
  .bounded_snapshot.checkpoint_attempt_count == 408 and
  .bounded_snapshot.post_boundary_candidate_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.post_boundary_ack_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$pair13"
jq -e '
  .attempt_count == 408 and
  .last_attempt.sequence == 407 and
  .policy_state.completed_pairs == 13 and
  .policy_state.next_sequence == 408 and
  (.policy_state.carriers | length) == 238 and
  ([.policy_state.carriers[] | select(. != 0)] | length) == 238 and
  ([.policy_state.carriers[]] | unique | length) == 231 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | unique | length) == 226 and
  ([.policy_state.carriers[] | select(. > 0)] | length) == 116 and
  ([.policy_state.carriers[] | select(. < 0)] | length) == 122 and
  ([.policy_state.carriers[]] | min) == -836592 and
  ([.policy_state.carriers[]] | max) == 656915 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | min) == 218 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | add) == 54354123 and
  ([.policy_state.carriers[]] | add) == 5 and
  .selector_generation == 2106 and
  .selector_order_sha256 ==
    "6e862112c25932cc69c8319637bb8cc17a4bb759c2f49b7d4e3936d5743e5ddf" and
  .last_attempt.selector_ack.selector_query_order_sha256 ==
    "2e9dfc74b6c11ce881d2684446b4aef1341bbf14bf5d76567381469332768778" and
  .target_or_reveal_consumed == false and
  .terminal == null
' "$state13"
test "$(jq -c -j '.policy_state.carriers' "$state13" | shasum -a 256 | cut -d ' ' -f 1)" = \
  09ec4c13006ccd286d23dd2d953a6673d14008e193bef562b3cd2b9365468663
jq -e '
  .classification == "MATCHED_ADAPTIVE_PREFIX" and
  .candidate_endpoint.receipt_count == 416 and
  .candidate_endpoint.status_counts["0"] == 416 and
  .candidate_endpoint.status_counts["10"] == 0 and
  .candidate_endpoint.status_counts["20"] == 0 and
  .candidate_endpoint.core_count == 0 and
  .candidate_endpoint.model_count == 0 and
  .candidate_endpoint.terminal == false and
  .matched_blocks[-1].sequence_start == 408 and
  .matched_blocks[-1].sequence_end == 415 and
  .matched_blocks[-1].candidate.totals.conflicts == 32774 and
  .matched_blocks[-1].candidate.totals.decisions == 47963 and
  .matched_blocks[-1].candidate.totals.propagations == 3417505255 and
  .matched_blocks[-1].control.totals.conflicts == 32771 and
  .matched_blocks[-1].control.totals.decisions == 58825 and
  .matched_blocks[-1].control.totals.propagations == 3400084400 and
  .matched_blocks[-1].route_pair == "NOT_GREEN|NOT_GREEN" and
  ((.matched_blocks[-1].candidate.totals.decisions -
    .matched_blocks[-1].control.totals.decisions) ==
    ([.matched_blocks[] | select(.sequence_start >= 200) |
      (.candidate.totals.decisions - .control.totals.decisions)] | min)) and
  (.aggregate.candidate.cumulative_r192_through_analysis_end.conflicts -
    .matched_blocks[0].candidate.totals.conflicts) == 884896 and
  (.aggregate.candidate.cumulative_r192_through_analysis_end.decisions -
    .matched_blocks[0].candidate.totals.decisions) == 1234366 and
  (.aggregate.candidate.cumulative_r192_through_analysis_end.propagations -
    .matched_blocks[0].candidate.totals.propagations) == 95142237446 and
  (.aggregate.control.cumulative_r192_through_analysis_end.conflicts -
    .matched_blocks[0].control.totals.conflicts) == 884863 and
  (.aggregate.control.cumulative_r192_through_analysis_end.decisions -
    .matched_blocks[0].control.totals.decisions) == 1189319 and
  (.aggregate.control.cumulative_r192_through_analysis_end.propagations -
    .matched_blocks[0].control.totals.propagations) == 97207490120 and
  .aggregate.cumulative_delta_candidate_minus_control.conflicts == 33 and
  .aggregate.cumulative_delta_candidate_minus_control.decisions == 45047 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -2065252674 and
  .candidate_policy_state.completed_pairs == 13 and
  .candidate_policy_state.current_pair_index == 13 and
  .candidate_policy_state.current_pair_minus_score ==
    "3610095713863275/48711134008" and
  .candidate_policy_state.current_pair_plus_score == null and
  .candidate_policy_state.last_policy_event.pair_complete == false and
  .candidate_policy_state.last_policy_event.polarity == -1 and
  .candidate_policy_state.carrier_summary.count == 238 and
  .candidate_policy_state.carrier_summary.nonzero == 238 and
  .candidate_policy_state.carrier_summary.minimum == -836592 and
  .candidate_policy_state.carrier_summary.maximum == 656915 and
  .candidate_policy_state.carrier_summary.l1 == 54354123 and
  .assumption_actuation.completed_actuation_proof_count == 224 and
  ([.assumption_actuation.selector_native_links[] |
    select(.sequence >= 408 and .sequence <= 415) |
    select(.actuation_proof_complete == true and
      .full_order_sha256 ==
        "6e862112c25932cc69c8319637bb8cc17a4bb759c2f49b7d4e3936d5743e5ddf" and
      .query_order_sha256 ==
        "2e9dfc74b6c11ce881d2684446b4aef1341bbf14bf5d76567381469332768778" and
      .query_order_count == 238)] | length) == 8 and
  .assumption_actuation.pending_next_actuation.sequence == 416 and
  .assumption_actuation.pending_next_actuation.selector_generation == 2114 and
  .assumption_actuation.pending_next_actuation.full_order_sha256 ==
    "ff31cceae3c914244f3092cc22229243a5b66a72f1ac20c8e99bc497b571ff61" and
  .assumption_actuation.pending_next_actuation.query_order_sha256 ==
    "b7389c8d2667d916d8bd02df5fde328eb4466ab3eb829433589cdd6b134bd128" and
  .boundary.next_fixed_boundary_sequence_end == 423 and
  .bounded_snapshot.mode == "copied-checkpoint-boundary-complete-lines" and
  .bounded_snapshot.retrospective_prefix == false and
  .bounded_snapshot.checkpoint_attempt_count == 416 and
  .bounded_snapshot.post_boundary_candidate_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.post_boundary_ack_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$field13"
jq -e '
  .classification == "MATCHED_ADAPTIVE_PREFIX" and
  .candidate_endpoint.receipt_count == 424 and
  .candidate_endpoint.status_counts["0"] == 424 and
  .candidate_endpoint.status_counts["10"] == 0 and
  .candidate_endpoint.status_counts["20"] == 0 and
  .candidate_endpoint.core_count == 0 and
  .candidate_endpoint.model_count == 0 and
  .candidate_endpoint.terminal == false and
  .matched_blocks[-1].sequence_start == 416 and
  .matched_blocks[-1].sequence_end == 423 and
  .matched_blocks[-1].candidate.totals.conflicts == 32772 and
  .matched_blocks[-1].candidate.totals.decisions == 54306 and
  .matched_blocks[-1].candidate.totals.propagations == 3251829561 and
  .matched_blocks[-1].control.totals.conflicts == 32774 and
  .matched_blocks[-1].control.totals.decisions == 57591 and
  .matched_blocks[-1].control.totals.propagations == 3411245392 and
  .matched_blocks[-1].route_pair == "NOT_GREEN|NOT_GREEN" and
  (([.matched_blocks[-2:][] | .candidate.totals.conflicts] | add) -
    ([.matched_blocks[-2:][] | .control.totals.conflicts] | add)) == 1 and
  (([.matched_blocks[-2:][] | .candidate.totals.decisions] | add) -
    ([.matched_blocks[-2:][] | .control.totals.decisions] | add)) == -14147 and
  (([.matched_blocks[-2:][] | .candidate.totals.propagations] | add) -
    ([.matched_blocks[-2:][] | .control.totals.propagations] | add)) == -141994976 and
  ([range(0; 14) as $i |
    {
      decisions:
        ((.matched_blocks[1 + ($i * 2)].candidate.totals.decisions +
          .matched_blocks[2 + ($i * 2)].candidate.totals.decisions) -
         (.matched_blocks[1 + ($i * 2)].control.totals.decisions +
          .matched_blocks[2 + ($i * 2)].control.totals.decisions)),
      propagations:
        ((.matched_blocks[1 + ($i * 2)].candidate.totals.propagations +
          .matched_blocks[2 + ($i * 2)].candidate.totals.propagations) -
         (.matched_blocks[1 + ($i * 2)].control.totals.propagations +
          .matched_blocks[2 + ($i * 2)].control.totals.propagations))
    } |
    select(.decisions < 0 and .propagations < 0) |
    .decisions] | min) == -14147 and
  ([range(0; 14) as $i |
    {
      decisions:
        ((.matched_blocks[1 + ($i * 2)].candidate.totals.decisions +
          .matched_blocks[2 + ($i * 2)].candidate.totals.decisions) -
         (.matched_blocks[1 + ($i * 2)].control.totals.decisions +
          .matched_blocks[2 + ($i * 2)].control.totals.decisions)),
      propagations:
        ((.matched_blocks[1 + ($i * 2)].candidate.totals.propagations +
          .matched_blocks[2 + ($i * 2)].candidate.totals.propagations) -
         (.matched_blocks[1 + ($i * 2)].control.totals.propagations +
          .matched_blocks[2 + ($i * 2)].control.totals.propagations))
    } |
    select(.decisions < 0 and .propagations < 0) |
    .propagations] | min) == -141994976 and
  (.aggregate.candidate.cumulative_r192_through_analysis_end.conflicts -
    .matched_blocks[0].candidate.totals.conflicts) == 917668 and
  (.aggregate.candidate.cumulative_r192_through_analysis_end.decisions -
    .matched_blocks[0].candidate.totals.decisions) == 1288672 and
  (.aggregate.candidate.cumulative_r192_through_analysis_end.propagations -
    .matched_blocks[0].candidate.totals.propagations) == 98394067007 and
  (.aggregate.control.cumulative_r192_through_analysis_end.conflicts -
    .matched_blocks[0].control.totals.conflicts) == 917637 and
  (.aggregate.control.cumulative_r192_through_analysis_end.decisions -
    .matched_blocks[0].control.totals.decisions) == 1246910 and
  (.aggregate.control.cumulative_r192_through_analysis_end.propagations -
    .matched_blocks[0].control.totals.propagations) == 100618735512 and
  .aggregate.cumulative_delta_candidate_minus_control.conflicts == 31 and
  .aggregate.cumulative_delta_candidate_minus_control.decisions == 41762 and
  .aggregate.cumulative_delta_candidate_minus_control.propagations == -2224668505 and
  .candidate_policy_state.completed_pairs == 14 and
  .candidate_policy_state.last_policy_event.pair_complete == true and
  .candidate_policy_state.last_policy_event.pair_index == 13 and
  .candidate_policy_state.last_policy_event.pair_minus_score ==
    "3610095713863275/48711134008" and
  .candidate_policy_state.last_policy_event.pair_plus_score ==
    "3375592803987/60668872" and
  .candidate_policy_state.last_policy_event.contrast ==
    "-1137322527930592141373/7988528920907726878452" and
  .candidate_policy_state.last_policy_event.contrast_ppm == -142369 and
  .assumption_actuation.completed_actuation_proof_count == 232 and
  .assumption_actuation.selector_native_link_count == 232 and
  ([.assumption_actuation.selector_native_links[] |
    select(.sequence >= 416 and .sequence <= 423) |
    select(.actuation_proof_complete == true and
      .full_order_sha256 ==
        "ff31cceae3c914244f3092cc22229243a5b66a72f1ac20c8e99bc497b571ff61" and
      .query_order_sha256 ==
        "b7389c8d2667d916d8bd02df5fde328eb4466ab3eb829433589cdd6b134bd128" and
      .query_order_count == 238)] | length) == 8 and
  .assumption_actuation.pending_next_actuation.sequence == 424 and
  .assumption_actuation.pending_next_actuation.selector_generation == 2122 and
  .assumption_actuation.pending_next_actuation.full_order_sha256 ==
    "c17a92bff40cd9ed2a97c7fe43eb4e6d2d21d819d9b3b50416f9b14938eee74a" and
  .assumption_actuation.pending_next_actuation.query_order_sha256 ==
    "1b7c1e68dc082282b13e69011782eb9dcbfd46ea51072e5552059a8f12a5e521" and
  .boundary.next_fixed_boundary_sequence_end == 431 and
  .bounded_snapshot.mode == "copied-checkpoint-boundary-complete-lines" and
  .bounded_snapshot.retrospective_prefix == false and
  .bounded_snapshot.checkpoint_attempt_count == 424 and
  .bounded_snapshot.post_boundary_candidate_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.post_boundary_ack_bytes_supplied_to_sealed_analyzer == 0 and
  .bounded_snapshot.runtime_files_written == false and
  .bounded_snapshot.process_signals_sent == false and
  .bounded_snapshot.target_or_reveal_read == false
' "$pair14"
jq -e '
  .attempt_count == 424 and
  .last_attempt.sequence == 423 and
  .policy_state.completed_pairs == 14 and
  .policy_state.next_sequence == 424 and
  (.policy_state.carriers | length) == 238 and
  ([.policy_state.carriers[] | select(. != 0)] | length) == 238 and
  ([.policy_state.carriers[]] | unique | length) == 237 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | unique | length) == 234 and
  ([.policy_state.carriers[] | select(. > 0)] | length) == 124 and
  ([.policy_state.carriers[] | select(. < 0)] | length) == 114 and
  ([.policy_state.carriers[]] | min) == -875013 and
  ([.policy_state.carriers[]] | max) == 778755 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | min) == 913 and
  ([.policy_state.carriers[] | if . < 0 then -. else . end] | add) == 61419400 and
  ([.policy_state.carriers[]] | add) == 2 and
  .selector_generation == 2122 and
  .selector_order_sha256 ==
    "c17a92bff40cd9ed2a97c7fe43eb4e6d2d21d819d9b3b50416f9b14938eee74a" and
  .last_attempt.selector_ack.selector_query_order_sha256 ==
    "1b7c1e68dc082282b13e69011782eb9dcbfd46ea51072e5552059a8f12a5e521" and
  .target_or_reveal_consumed == false and
  .terminal == null
' "$state14"
test "$(jq -c -j '.policy_state.carriers' "$state14" | shasum -a 256 | cut -d ' ' -f 1)" = \
  343d428fedb88e25c96f7b6b3f73fe5210f2cface3f8f8e157ae8c7deee0c915
jq -n -e \
  --slurpfile current "$state14" \
  --slurpfile prior "$state13" '
  ($current[0].policy_state.carriers) as $c |
  ($prior[0].policy_state.carriers) as $p |
  ([range(0; 238) | select($c[.] != $p[.])] | length) == 238 and
  ([range(0; 238) |
    select((($c[.] > 0) and ($p[.] < 0)) or
           (($c[.] < 0) and ($p[.] > 0)))] | length) == 38
'
jq -n -e \
  --slurpfile current "$pair14" \
  --slurpfile prior_max "$field11" '
  $current[0].aggregate.cumulative_delta_candidate_minus_control.propagations ==
    -2224668505 and
  $prior_max[0].aggregate.cumulative_delta_candidate_minus_control.propagations ==
    -2183021242 and
  ((-$current[0].aggregate.cumulative_delta_candidate_minus_control.propagations) -
   (-$prior_max[0].aggregate.cumulative_delta_candidate_minus_control.propagations)) ==
    41647263
'

# Portable package smoke paths
.venv/bin/o1-crypto-lab benchmark \
  --config configs/quick.json \
  --output runs/quick.json
.venv/bin/o1-crypto-lab living-inverse-foundation \
  --config configs/living_inverse_foundation_v1.json
.venv/bin/o1-crypto-lab full256-paired-sensor \
  --config configs/full256_paired_causal_sensor_v1.json

Install -e '.[train]' instead of -e . for the historical learned-O1 tests. Some archived experiments deliberately bind an exact CPython build and source closure; their freeze tests are provenance checks, not portable smoke tests. In particular, test_o1c0237_launch.py requires the exact local O1C-0213 terminal LIVE transcript used by the sealed preflight. That mutable-source dependency is intentionally not reconstructed or silently substituted in a fresh clone. The current live O1C-0237 identity must not be relaunched or replayed while it is active. Its controller, policy, analyzer and fixed artifacts are public; the mutable runtime becomes a frozen terminal package only after natural completion.

Inspect the operator compiler and its leakage rejection:

.venv/bin/o1-crypto-lab compose
.venv/bin/o1-crypto-lab boundary

Replay the supplied real O1-O session as normalized evidence without importing or executing any generated program. This optional command expects a local checkout of DT-Foss/O1-O; substitute its real path:

.venv/bin/o1-crypto-lab replay-o1o \
  --session /path/to/O1-O/2026-02-18_013412 \
  --output runs/o1o-2026-02-18-replay.json \
  --include-events

The replay intentionally distinguishes generation success, process success, capability evidence and mission progress. Raw stdout/stderr is never copied into the O1 stream; only its length and an unsalted integrity fingerprint remain until a domain-specific parser has validated its semantics. That fingerprint is not a confidentiality guarantee for guessable output. engagement_report.json contributes hashed aggregate retry/recovery/chaining counts, and every normalized event is fed neutrally into a capped TargetModel without being mislabeled as research success. The February format did not retain explicit retry-parent IDs, so the replay reports aggregate topology rather than inventing edges.

Verify a published snapshot before an adapter reads it:

.venv/bin/o1-crypto-lab verify-source \
  --root ../fullround-key-recovery \
  --manifest ../fullround-key-recovery/provenance/ARTIFACTS.sha256 \
  --output runs/fullround-source-verification.json

Run the manifest-pinned Stage-3 ingestion and frozen retrospective reader protocol:

.venv/bin/o1-crypto-lab stage3-ingest \
  --config configs/stage3_a296_a297_ingest_v1.json
.venv/bin/o1-crypto-lab stage3-reader \
  --config configs/stage3_reader_retrospective_v1.json

Each command creates a new read-only directory named runs/YYYYMMDD_HHMMSS_O1C-.../ with RUN.md, exact config and command, environment, metrics, logs, checkpoints, retained artifacts and a complete SHA-256 manifest. Attempt IDs are permanently reserved and cannot overwrite a prior result. Verify any capsule with:

.venv/bin/o1-crypto-lab verify-run runs/<capsule-name>

Recompute the label-using O1C-0018 diagnostics without generating a target or changing any model state:

PYTHONPATH=src .venv/bin/python \
  -m o1_crypto_lab.full256_online_real_forensics \
  --capsule runs/20260717_152827_O1C-0018_full256-online-real-gate-dev-v1

The Direct12 dependency chain originally lived in a dirty sibling worktree. It is therefore curated honestly into its own immutable capsule instead of being called a clean Fullround-manifest artifact:

.venv/bin/o1-crypto-lab direct12-snapshot \
  --config configs/direct12_source_snapshot_v1.json

Reproduce the frozen 133-to-532 trajectory reader from that capsule, then run the bounded-memory mechanism tournament:

.venv/bin/o1-crypto-lab direct12-reproduce \
  --config configs/direct12_reproduction_v1.json
.venv/bin/o1-crypto-lab bounded-memory-tournament \
  --config configs/bounded_memory_tournament_v1.json
.venv/bin/o1-crypto-lab corrected-codec-bridge \
  --config configs/corrected_codec_bridge_v1.json
.venv/bin/o1-crypto-lab upstream-ising-freeze \
  --config configs/upstream_ising_retrospective_v1.json

The tournament compares global Walsh state, sixteen fixed low4/high8 slot banks, and a dense 2–8-bit integer Bit-Vault. O1-O sees only A348 target-blind fidelity, serialized online-state bytes, update work and clip counts. It persists one future template before the A349 score member is opened; all complete A349 orders are then persisted before the separate A348 truth API is called. Direct candidate tables and full spectral banks remain clearly labeled ceilings and cannot win the mechanism gate. The dense Bit-Vault is a full-rank 4,080-register mechanism, not a claim of sublinear capacity.

The corrected-codec bridge reproduces A355/A356 exactly and retains the selected 6-bit DC-complete bank only as a validation ceiling. Its 4,096 spectral degrees of freedom are information-equivalent to the fixed candidate table, and its 8,014-byte maximum serialized logical state is larger than the matched 3,918-byte direct baseline.

The upstream freeze replaced the dense final-field representation with a 12-register unary solver-evidence memory. Its conservative logical-state bound is 266 bytes and its frozen binary is 162 bytes. The complete target-blind A355 panel contained 672 orders; the frozen decoder ranked the retrospective target at 73, but the exact conditional random-label tail was 2431/4096 = 0.593505859375. That is a structurally eligible compact mechanism and a negative efficacy result, not SOTA. The same decoder emitted a complete A356 order before any A356 target or outcome read, but A356 still came from the same opened source capsule and is not a source-unseen holdout. That planned narrow W46 follow-up is now superseded. O1C-0008 instead freezes the full-256 public-output attacker type, separately typed teacher labels, exact traced relation generator, six Contrast-Key families, sealed full-256 broker and the complete non-recovery progress vector. The unary decoder remains one matched baseline inside the new architecture.

The expensive immutable-snapshot integration gates are opt-in:

O1_CRYPTO_DIRECT12_REAL=1 \
  .venv/bin/python -m unittest discover -s tests -v
O1_CRYPTO_CORRECTED_REAL=1 \
  .venv/bin/python -m unittest discover -s tests -v
O1_CRYPTO_UPSTREAM_REAL=1 \
  .venv/bin/python -m unittest discover -s tests -v

Integration contract

No imports from the dirty live O1, O1-O, or arx-carry-leak trees.
Packaged benchmark bytes are confined to runs/. Native resident attempts use one isolated attempt-local scratch/<attempt>/runtime/ directory. Neither path can write into a sibling project. Mutable live runtime is excluded from commits; fixed boundary artifacts and terminal packages are added explicitly.
A separate symlink-safe writer may update only the enumerated cockpit Markdown files at the lab root and under research/; it cannot write arbitrary lab or sibling paths.
CPU-only by default. MPS/GPU use requires a short explicit resource-checked window and cannot compete with the active sibling recovery queue.
No target label, recovered model, post-reveal rank, or target-internal state may flow into a TARGET_BLIND_ORDER.
Training-time internal round states may teach an operator, but evaluation-time features must be public or recomputable from public equations under an explicitly billed candidate assumption.
Every learned operator is selected on training/validation keys, frozen and hashed, then evaluated once on disjoint test keys with matched controls. The lifecycle is enforced as DISCOVERY -> FROZEN -> TEST_CONSUMED -> AUDIT; the exact proposal and plan hashes are bound at freeze time.
State size, external-index growth, samples and total cipher/solver work are reported separately. O(1) always means constant in stream length unless another axis is named explicitly.

See Architecture, O1-256 Living Inverse, Experiment ladder, and Scientific boundaries. The first reproducible smoke-test measurements are recorded in First results.

Repository map

configs/                 deterministic benchmark configurations
docs/                    architecture, gates and claim boundaries
provenance/              lab-owned byte ledgers and source-boundary records
research/                hypotheses, append-only attempts, breadcrumbs, next actions
src/o1_crypto_lab/       memory, composer, replay, TargetModel, adapters and CLI
tests/                   unit, leakage and reproducibility gates
runs/                    immutable timestamped capsules (ignored by Git, never overwritten)
scratch/                 native engines, attempt-local policies and frozen boundary artifacts

The design is derived from the Apache-2.0 O1 and O1-O projects. No source from their offensive modules is executed or imported here; only the domain-agnostic bounded-state, typed-composition, verification and failure-memory ideas are used.

Name		Name	Last commit message	Last commit date
Latest commit History 621 Commits
configs		configs
docs		docs
native		native
provenance		provenance
research		research
runs		runs
scratch		scratch
src/o1_crypto_lab		src/o1_crypto_lab
tests		tests
.gitignore		.gitignore
CITATION.cff		CITATION.cff
GOAL.md		GOAL.md
LICENSE		LICENSE
NOTICE		NOTICE
README.md		README.md
RELEASE_NOTES.md		RELEASE_NOTES.md
REPRODUCIBILITY.md		REPRODUCIBILITY.md
RESULTS_INDEX.md		RESULTS_INDEX.md
STATUS.md		STATUS.md
pyproject.toml		pyproject.toml
pytest.ini		pytest.ini
requirements-dev.txt		requirements-dev.txt

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

Cryptanalytic Memory Lab (`cryptO1`)

Current frontier — 2026-07-25

Proven mechanism and exact-search ladder

O1C-0237: adaptive bounded-state result

Architecture

Experimental contract

Active Moonshot: O1-256 Living Inverse

Public disclosure and claim boundary

What the first benchmark proves

Run

Integration contract

Repository map

About

Releases

Packages

Contributors

Languages

Folders and files

Latest commit

History

Repository files navigation

Cryptanalytic Memory Lab (cryptO1)

Current frontier — 2026-07-25

Proven mechanism and exact-search ladder

O1C-0237: adaptive bounded-state result

Architecture

Experimental contract

Active Moonshot: O1-256 Living Inverse

Public disclosure and claim boundary

What the first benchmark proves

Run

Integration contract

Repository map

About

Topics

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages

Cryptanalytic Memory Lab (`cryptO1`)