1 · STATE OF THE BOOK

COLDREAD · LIVE SOL BOOK · THREE SLEEVES · opened 2026-08-09 08:51Z · 1:51 AM · paper
10.19 SOL-168.15 SOL
LIVE SOL BOOK (cash + every marked open position, three sleeves) · SOL phase 2 confidence-weighted-fresh-sol-paper since 2026-08-13 00:28Z · 5:28 PM · vs 28.34 SOL phase baseline · 3600 phase closed / 5538 all-time closed · 7 open · ledger 1m old
Headline is complete-ledger cash plus marked normal and displaced positions, minus the phase baseline and phase funding. All-time exact realized is -240.09 SOL. Not converted to USD: a SOL P&L times the CURRENT rate is not the sum of each trade at its own rate.
-169.05 SOLphase realized SOL
+150.00 SOLphase funding
9.57 SOLcash bankroll
0.62 SOLmarked open value
3600closed trades in phase
7open positions (three sleeves)
0displaced funded positions
5545buys in window
5537sells in window

Performance & trend

3600 exact phase closes · 3599 chartable tail rows
SOL phase 2 confidence-weighted-fresh-sol-paper since 2026-08-13 00:28Z · 5:28 PM · 1938 older closed trips excluded from this current-performance view.
-169.05 SOLcurrent-phase realized (3600 exact closes)
-185.15 SOLcurrent-phase ex-top-1
9.57 SOLcurrent bankroll
+16.97 SOLbest current-phase trade

Current-phase P&L path

Realized P&L at each closed-trip checkpoint; the latest point is funding-adjusted total-book P&L and includes all marked open positions. Endpoint = the headline total above.
max +0.00 SOL min -168.18 SOL 2026-08-13 2026-09-02

Current-phase trend

Last 25 closed trips vs the 25 before them, by trip order (not calendar). Windows stated so the comparison cannot be re-cut after the fact.
windowrealizedex-top-1span
last 25-0.45 SOL-0.51 SOL2026-09-02 01:48Z · 6:48 PM → 2026-09-02 19:45Z · 12:45 PM
prior 25-0.49 SOL-0.52 SOL2026-09-01 08:51Z · 1:51 AM → 2026-09-02 01:28Z · 6:28 PM
IMPROVING +0.04 SOL on the total, +0.01 SOL ex-top-1.
Both windows are negative. The direction above describes which loss is smaller, not a book that is making money. Ex-top-1 is printed for each window because one trade moves a 25-trip total on this book. No significance test is claimed — this is a description of two windows.

Kill criteria

K1 FIRED K2 NOT-FIRED K3 UNEVALUABLE K4 UNEVALUABLE K5 FIRED RK1 ARMED · NOT-YET
KILL BAR FIRED: K1, K5 — pre-registered stop verdicts standing as-of 13m ago (2026-09-02 20:12Z). The K-arm is operator-STOPPED (context below); the fire stays loud until re-baseline or re-registration.
RK1 forward clock (runner-v1): 156/200 forward entries · baseline 501 · armed 4.8d ago (2026-08-29 01:10Z) by Jeremy (operator)
K1–K5 describe the operator-STOPPED live arm (Arm 1 (live, birth-entry) stopped by operator ruling, 2026-08-28 ~1:30 PM MST (2026-08-28T20:30Z)) — a standing FIRED there is the incident that stop answered. RK1 binds runner-v1, which is running.
as-of k1Watch 13m ago (2026-09-02 20:12Z) · rk1Watch 8m ago (2026-09-02 20:18Z) · 2138 arm entries

Open positions

book summary

7 open · 0.64 SOL deployed
Live SOL book, replayed from the ledger. Individual positions and their latest executable marks appear once in the table below. No SOL/USD rate is applied. · last ledger row 1m ago (2026-09-02 20:24Z · 1:24 PM)

positions

count and deployed SOL are the book-summary line above — stated once. Live SOL book, replayed from the complete shared ledger (three sleeves); PNL below is MARKED against current executable state, entry/size/tokens are not
positionsizeheld for~entry px (SOL/tok)~tokenspnlpnl %
Booby SOLTHESIS UNKNOWNmeasurement-missing:markVsEntry · measurement-missing:quoteReserveVsEntryCA GDTj9…zpump
details
pool 6LxgMG8dS5… · name metadata: pump-creates.log.jsonl @ 2026-09-02T08:02:31.284Z
Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
thesis: measurement-missing:markVsEntry · measurement-missing:quoteReserveVsEntry
 0.10 SOL 11h 12m ~0.00000002895 ~3.44M ~-0.0014 SOL -1.4% CA GDTj9…zpump
details
pool 6LxgMG8dS5… · name metadata: pump-creates.log.jsonl @ 2026-09-02T08:02:31.284Z
Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
thesis: measurement-missing:markVsEntry · measurement-missing:quoteReserveVsEntry
pippo SOLTHESIS UNKNOWNmeasurement-missing:markVsEntry · measurement-missing:quoteReserveVsEntryCA FDaGG…vaiqB
details
pool 6ekyJYCYb5… · name metadata: pump-creates.log.jsonl @ 2026-09-02T12:13:53.683Z
Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
thesis: measurement-missing:markVsEntry · measurement-missing:quoteReserveVsEntry
thesis grade: VALIDATED_BUT_RED* · post-verdict -0.3% ?
 0.10 SOL 7h 2m ~0.00000002445 ~3.99M ~-0.0018 SOL -1.8% CA FDaGG…vaiqB
details
pool 6ekyJYCYb5… · name metadata: pump-creates.log.jsonl @ 2026-09-02T12:13:53.683Z
Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
thesis: measurement-missing:markVsEntry · measurement-missing:quoteReserveVsEntry
thesis grade: VALIDATED_BUT_RED* · post-verdict -0.3% ?
FAMI SOLTHESIS INVALIDATEDfalsifier:tape.prints.last300sCA 92brN…2in5u
details
pool 8ZhfkKVrKB… · name metadata: pump-creates.log.jsonl @ 2026-09-02T14:23:09.791Z
Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
thesis: falsifier:tape.prints.last300s
thesis grade: VALIDATED_BUT_RED* · post-verdict -3.6% ?
 0.10 SOL 4h 49m ~0.00000002925 ~3.37M ~-0.0049 SOL -5.0% CA 92brN…2in5u
details
pool 8ZhfkKVrKB… · name metadata: pump-creates.log.jsonl @ 2026-09-02T14:23:09.791Z
Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
thesis: falsifier:tape.prints.last300s
thesis grade: VALIDATED_BUT_RED* · post-verdict -3.6% ?
Bros SOLTHESIS UNKNOWNmeasurement-missing:markVsEntry · measurement-missing:quoteReserveVsEntryCA 75etQ…Dpump
details
pool EL5iTE8b6c… · name metadata: pump-creates.log.jsonl @ 2026-09-02T14:45:37.871Z
Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
thesis: measurement-missing:markVsEntry · measurement-missing:quoteReserveVsEntry
 0.06 SOL 4h 10m ~0.00000001875 ~3.13M ~-0.00066 SOL -1.1% CA 75etQ…Dpump
details
pool EL5iTE8b6c… · name metadata: pump-creates.log.jsonl @ 2026-09-02T14:45:37.871Z
Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
thesis: measurement-missing:markVsEntry · measurement-missing:quoteReserveVsEntry
$LEMSICK SOLTHESIS INVALIDATEDfalsifier:tape.prints.last300sCA 2YAY8…Lpump
details
pool 6Eoaw4bU7c… · name metadata: pump-creates.log.jsonl @ 2026-09-02T15:35:50.674Z
Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
thesis: falsifier:tape.prints.last300s
thesis grade: VALIDATED_BUT_RED* · post-verdict -9.6% ?
 0.10 SOL 3h 10m ~0.0000004006 ~241,618 -0.01 SOL -8.0% CA 2YAY8…Lpump
details
pool 6Eoaw4bU7c… · name metadata: pump-creates.log.jsonl @ 2026-09-02T15:35:50.674Z
Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
thesis: falsifier:tape.prints.last300s
thesis grade: VALIDATED_BUT_RED* · post-verdict -9.6% ?
UNDEADS SOLTHESIS UNKNOWNmeasurement-missing:markVsEntry · measurement-missing:quoteReserveVsEntryCA EguEN…xpump
details
pool CxQYptHP5B… · name metadata: pump-creates.log.jsonl @ 2026-09-02T16:35:59.438Z
Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
thesis: measurement-missing:markVsEntry · measurement-missing:quoteReserveVsEntry
thesis grade: VALIDATED_BUT_RED* · post-verdict -1.7% ?
 0.10 SOL 3h 3m ~0.00000003623 ~2.64M ~-0.0029 SOL -3.0% CA EguEN…xpump
details
pool CxQYptHP5B… · name metadata: pump-creates.log.jsonl @ 2026-09-02T16:35:59.438Z
Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
thesis: measurement-missing:markVsEntry · measurement-missing:quoteReserveVsEntry
thesis grade: VALIDATED_BUT_RED* · post-verdict -1.7% ?
Hero SOLTHESIS SUPPORTEDCA Ck377…Npump
details
pool FXLxdZyX2Z… · name metadata: pump-creates.log.jsonl @ 2026-09-02T19:37:42.347Z
Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
thesis: all-confirmations-met
 0.10 SOL 2m ~0.0000002691 ~359,240 -0.01 SOL -6.3% CA Ck377…Npump
details
pool FXLxdZyX2Z… · name metadata: pump-creates.log.jsonl @ 2026-09-02T19:37:42.347Z
Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
thesis: all-confirmations-met

Recent execution & attempts
Newest 8 events in the selected window.
EXECUTED is a paper fill; other states are attempts/blocks, not fills. Full history and row detail in Review.
09-02
20:24Z · 1:24 PMHero SOL Ck377NeTAQ1ayhBtNjm3umscNkpekBtcJHjPJrtNpump BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
09-02
19:45Z · 12:45 PMHero SOL Ck377NeTAQ1ayhBtNjm3umscNkpekBtcJHjPJrtNpump SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · trail
09-02
19:42Z · 12:42 PMHero SOL Ck377NeTAQ1ayhBtNjm3umscNkpekBtcJHjPJrtNpump BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
09-02
19:06Z · 12:06 PM8Bit SOL 4XbRwc96Bd4QbMMepT27Naz95GFUCoXd7KoUfgy9pump SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop
09-02
18:45Z · 11:45 AMDEG SOL 7HJCxCMTnit9rLS21e9SMdisJaaZ9A5P3N1Lq99roeTz SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · trail
09-02
17:23Z · 10:23 AMUNDEADS SOL EguENXBejC9EvHqqfKZ1koVnUNbbYM1EcLsPcJtxpump BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
09-02
17:16Z · 10:16 AM$LEMSICK SOL 2YAY8PucaY9zVFvyUArjMB12HxtRRrrTtzLRuwpLpump BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
09-02
17:12Z · 10:12 AMDEG SOL 7HJCxCMTnit9rLS21e9SMdisJaaZ9A5P3N1Lq99roeTz BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner
2 · GO / NO-GO

Readiness trajectory

RECEIPT V1
GATE 0 / 7 INSUFFICIENT
Sequential, non-compensating readiness · receipt rrv1_2c85bfeed4907e1… · COMMITTED HEAD · 2026-09-02 14:50:29.568Z · PARTIAL
Authorizations: pilot not authorized · scaling not authorized (config/readiness-authorizations.v1.json absent: nothing is authorized) · 0 check(s) not applicable until their trigger
Separate provisional overlays: DIRTY WORKTREE OBSERVATION 2026-08-29 02:23:27.298Z (REGRESSING; never committed history)
External evidence ledger: 33 validated attestation row(s) · latest attestation joined to 25 check id(s).
G0Architecture integrity6/6 ↑ +1 vs priorVERIFIED
VERIFIED declared-runtime-book · owner ARCHITECTURE · decays in 24h
exit 0 in 28268ms: TAP version 13 # Subtest: retired legacy pre-entry replay cannot regain measurement authority ok 1 - retired legacy pre-entry replay cannot regain measurement authority ---…
WHAT'S NEEDED: retain fresh receipt
VERIFIED current-target-maps-and-ois · owner ARCHITECTURE · decays in 24h
exit 0 in 2657ms: { "status": "OK", "undocumented": [], "undocumentedTotal": 57, "liveModules": 475, "entrypoints": 94, "closureTruncated": false, "baselineDate": "2026-07-24", "baselineIntact": true,…
WHAT'S NEEDED: retain fresh receipt
VERIFIED source-typecheck · owner CODEX · decays in 24h
exit 0 in 21953ms
WHAT'S NEEDED: retain fresh receipt
VERIFIED test-typecheck · owner CODEX · decays in 24h
exit 0 in 37992ms
WHAT'S NEEDED: retain fresh receipt
VERIFIED terminating-fast-test-suite · owner CODEX · decays in 24h
exit 0 in 457131ms: [buildTestCache] Pre-compiling and bundling 965 test suites with esbuild... ✔ active reserve marker validates pool and vault mints, then returns an executable RPC-vault mark…
WHAT'S NEEDED: retain fresh receipt
VERIFIED clean-canonical-node22-release · owner ARCHITECTURE · decays in 24h
exit 2 in 3391ms: { "schemaVersion": "coldread-environment-doctor-v1", "readOnly": true, "startedAt": "2026-09-02T14:59:39.374Z", "finishedAt": "2026-09-02T14:59:41.365Z", "durationMs": 1991.6081,…
WHAT'S NEEDED: retain fresh receipt
G1Data integrity3/5 → flat vs priorREFUSE
VERIFIED immutable-data-identity · owner DATA · decays in 21h
Executed double-pass content-address audit at HEAD 1e1ff91d3e8e: 11 append-only canonical receipt tapes under data/receipts (86,273,073 bytes of captured extent, 172,528,850 bytes digested twice) plus…
WHAT'S NEEDED: re-run node --import tsx src/cli/immutableDataIdentityAuditCli.ts --write --candidate from the repo root on this machine before the TTL and import the candidate with readinessExternalEvidenceCli --candidate <path> --write
REFUSE source-event-ingest-clocks · owner DATA · decays in 16h
bounded-tail-clock-discipline-audit g1seic_a9d84fe4033125f466f094d5 over 7 canonical event tapes (39324 rows in 2097152-byte tail windows, newest end, current generations only; 0 parse failures): FAIL…
WHAT'S NEEDED: fix the writers of the failing stream(s) (wide-runners.log.jsonl) so every row carries a valid source/event clock, then re-run --write --candidate and import
VERIFIED replayable-decision-time-universe · owner RESEARCH · decays in 21h
Executed bounded replay g1urr_cd91944750a2f0e8f811e7f4 over 2026-09-01T11:23:32.665Z..2026-09-02T11:23:32.666Z (24h, ledgers sol-universe-source-observations.jsonl, sol-universe-candidates.jsonl,…
WHAT'S NEEDED: re-run node --import tsx src/cli/universeReplayAuditCli.ts --write --candidate before the TTL and import the candidate with readinessExternalEvidenceCli --candidate <path> --write
REFUSE quantified-completeness-and-staleness · owner DATA · decays in 23h
Bounded rotation-aware coverage audit (source receipt g1cov_7cd19a7e0bee1245f7f2b6c9) REFUSED on 2 of 12 feeds: sol-tape.pump.jsonl: newest event 35183.329s old (over 3600s) | rh-curve.jsonl: newest…
WHAT'S NEEDED: resolve the refused feed(s) (sol-tape.pump.jsonl: newest event 35183.329s old (over 3600s) | rh-curve.jsonl: newest event 10264.396s old (over 3600s)) then re-run --write --candidate and import
VERIFIED leakage-and-survivorship-controls · owner RESEARCH · decays in 85h
Executed validation run g1lsv1_89e88689760bc19c at 2026-08-31T03:11:37.576Z: 11 enumerated leakage/survivorship controls, 10 PASS / 0 FAIL / 1 not-demonstrable. Preregistered clocks on all three…
WHAT'S NEEDED: Re-run the g1 leakage-survivorship validation audit after the four-action holdout opens (2026-08-30T04:03:56.275Z); the holdout-untouched control must then be re-proven from the selector's own holdout coverage counters instead of temporal impossibility, and a negative control for the non-causal-clock exclusion counter should be added.
G2Simulation fidelity0/3 → flat vs priorREFUSE
STALE decision-quote-versus-executable-fill · owner EXECUTION · TTL expired at receipt time
Executed bounded-tail execution-lineage audit at HEAD 91ce0f9f4096: 501 fills (250 entries, 251 exits, arms {"runner-v1":121,"live":380}) in the newest 16MiB of the live shared SOL book…
WHAT'S NEEDED: re-run the audit script embedded verbatim in the source receipt (scriptSource; sha256 recorded) from the repo root on this machine to produce a fresh source receipt and candidate, preview it with readinessExternalEvidenceCli --candidate, and hand the operator the --write import before the 24h TTL expires
REFUSE venue-reserve-route-cost-latency-failure-impact · owner EXECUTION · TTL expired at receipt time
bounded execution-context audit g2ecx_c67d5280fed910b264d46cd5 over the live SOL paper book tail (32MiB of data/arm1-engine-ledger.jsonl: 23234 rows, 1242 entry/exit, 1242 instrumented with…
WHAT'S NEEDED: build the two missing models and re-audit: (a) a VERSIONED venue-latency artifact the fill path can cite (chain-observed PumpSwap swap-landing sampler within Triton budget, or the cheaper versioned assumed-latency spec re-pricing fills at reserves observed T later) and (b) a VERSIONED venue failure-rate artifact (bounded failed-vs-landed swap share on comparable pools); optionally add the versioned lamport network/priority-fee term; then re-run this audit and re-import
REFUSE adverse-execution-sensitivity · owner RESEARCH · decays in 63h
Executed adverse-execution stress audit at HEAD 91ce0f9f4096 over the live OI-126 stress tape (bounded newest-end tail, 12,582,912 of 60,672,715 bytes): 4345 stress rows folded, all under the single…
WHAT'S NEEDED: To turn this green, in order: (a) obtain the operator's preregistered numeric viability bar (net expectancy after stress vs the solBenchmark null vs zero) via a COORDINATION.md ruling BEFORE any holdout verdict is visible; (b) register additional adverse axes (Vq tightening, added-latency re-pricing, failure/retry injection, deeper depth haircuts) as forward-only specs and let forward accrual cover them; (c) re-author after Gate 3 positive-net-expectancy turns — a receipt-backed viable baseline is a precondition. Meanwhile re-run the audit script embedded verbatim in the source receipt (scriptSource; sha256 recorded) from the repo root on this machine before expiry, preview with readinessExternalEvidenceCli --candidate, and hand the operator the --write import.
G3Strategy statistical validity1/5 → flat vs priorREFUSE
REFUSE purged-walk-forward · owner RESEARCH · decays in 63h
selector-receipt audit g3pwf_5d8ec237728329bb0dec0f7e over the live OI-126 selection instrument (data/sol-strategy-selection.json generated 2026-08-29T05:01:15.950Z, status REFUSE, 31 candidates = 30…
WHAT'S NEEDED: operator activates the prepared exit-policy successor v3 remint (sol-phase-2-exit-policy-successor-v3-20260826; 7d train window 604,800,000ms clears the 129,600,000ms minimum), then once >=30 settled train trips with >=10 retained per fold post-purge accrue (plausibly 2026-09-01/02, before its holdout opens 2026-09-02T15:00Z) re-run solStrategySelectionCli and re-author this check from the fresh receipt
UNKNOWN untouched-holdout · owner RESEARCH · TTL expired at receipt time
no evidence imported yet — produce a fresh identity-bound selector-receipt
VERIFIED multiplicity-and-outlier-control · owner RESEARCH · decays in 63h
Executed multiplicity+outlier-control audit at HEAD 91ce0f9f4096 (script + measurements frozen in the source receipt). (1) Live exit-policy selector receipt data/sol-strategy-selection.json (generated…
WHAT'S NEEDED: re-run the audit script embedded verbatim in the source receipt (scriptSource; sha256 recorded) from the repo root on this machine to produce a fresh source receipt and candidate, preview it with readinessExternalEvidenceCli --candidate, and hand the operator the --write import before expiry
UNKNOWN effective-sample-across-regimes · owner RESEARCH · decays in 88h
no evidence imported yet — Re-author the audit so `state` DERIVES from the computed verdict it already produces: gate on the L126 rederivedVerdict rather than passing a literal to the builder. Until then this check cannot supply evidence in either direction. The MARKETDATA lane is building a guard, anchored on `state` and reading both scriptSource and auditScript.source, that fails an audit which cannot express a PASS *or* one that cannot express a FAIL — this row should stay UNKNOWN until that guard passes the re-authored audit.
REFUSE positive-net-expectancy-after-costs · owner RESEARCH · decays in 63h
selector-receipt audit g3pnec_60c39cdf92844b1c58e705b1: REFUSE — the check's claim (the SELECTED strategy's holdout expectancy stays positive after realistic costs and under declared adverse…
WHAT'S NEEDED: pre-register the null bar for this check (positive net expectancy must beat the same-window SOL drift, not just zero) as a COORDINATION.md ruling BEFORE the four-action holdout opens 2026-08-30T04:03:56.275Z; then re-author from the first matured untouched-holdout receipt (four-action verdict due 2026-09-06T04:03:56.275Z, V5 2026-09-08T14:30:00Z) — green requires an ELIGIBLE challenger surviving positive holdout mean at every declared haircut level and under the preregistered stress
G4Risk validation0/3 → flat vs priorREFUSE
REFUSE independent-fail-closed-pretrade-risk · owner RISK · TTL expired at receipt time
executed audit g4ifc_a6a6946d76d376ed91395223b61d1eb1 (repo HEAD 91ce0f9f4, this box): REFUSE — no independent fail-closed pre-trade risk policy exists on the live path; both load-bearing properties…
WHAT'S NEEDED: wire an enforced independent pre-trade check into arm1EngineCli / arm1EstablishedEngineCli / arm1RunnerEngineCli — consume solPortfolioRiskAuthority behind solPortfolioRiskEnvelope, or promote preEntryFilterCore's FAIL_CLOSED dispositions from shadow to enforced — with a new refusal reasonCode written to the causal stream, stage declaration in architectureStages.ts in the same commit, hermetic tests proving both directions (blocks on null inputs, admits on complete-and-clean), then accrue forward receipts showing the block path firing on real candidates before any VERIFIED re-authoring; enforcement changes the entered population and stays operator sign-off
REFUSE tail-rug-liquidity-capacity-limits · owner RISK · decays in 63h
Executed risk-limit stress audit g4trlc_e6af5bf98a85692d0f99cd36 at HEAD 91ce0f9f4096: adversarial vector matrix through the exact exported limit cores the engines import, plus a bounded newest-end 16…
WHAT'S NEEDED: to turn this green, in order: (a) wire an enforced aggregate tail limit — consume solPortfolioRiskAuthority behind its envelope in all three engine CLIs (operator policy review required per RISK_SYSTEM.md) or have RK1-FIRED write the stop marker itself (operator ruling required on whether pager-plus-operator-halt suffices); (b) close the rug fail-open: flip unevaluable-veto disposition to fail-closed or promote the anti-rug funnel through its forward-validation path (both change the entered population — operator sign-off under experiment-autopilot governance); (c) promote the impact gate via scripts/entry-signal-autotune.mjs forward validation and fix its null-verdict fail-open; (d) build the failed-exit bound (assess src/cli/heldExitQuoteCli.ts reuse before building anything new); then re-run the audit script embedded verbatim in the source receipt (scriptSource, sha256 recorded) from the repo root on this machine and hand the operator the --write import. Meanwhile re-author before the 7d TTL expires.
REFUSE kill-switch-and-reconciliation-injection · owner RISK · decays in 63h
sandboxed failure-injection audit g4ksri_b0b39c5205e0fa3904f96ff6 (synthetic data ONLY in the scratchpad sandbox; production files read-only): KILL PATH — the real runnerKillCriteriaCli FIRED on a…
WHAT'S NEEDED: Level 2: add a --data-dir/--drill passthrough to scripts/rk1Watch.mjs (state file + board post redirected to the sandbox) and drive a synthetic FIRED through CLI -> rk1Watch asserting the 0/1/2/3 exit map; Level 3: watchdog -Drill mode routing Send-BankrAlert to a drill log (edit only at a recycle boundary — live-wrapper landmine) and traverse watchdog -> page under injection in an operator-approved window; fix the K5 fail-open (runnerKillCriteriaCli should refuse torn/malformed ledger lines, not skip them); obtain the operator ruling on auto-halt (does RK1-FIRED write the stop marker itself, or is pager+operator the accepted kill switch); then re-run this audit end-to-end and re-import within the 7d TTL
G5Shadow live-data validation0/3 → flat vs priorREFUSE
REFUSE sustained-zero-capital-shadow · owner OPERATIONS · TTL expired at receipt time
runtime shadow-desk audit g5szcs_90de25256453e37a8d6d4e60dcc5986d executed 2026-08-29T06:05:53.689Z on this box: REFUSE — the RUNTIME_DEPLOYED claim (sustained real-time shadow orders with no capital,…
WHAT'S NEEDED: configure COLDREAD_RELEASE_EVIDENCE_ROOT outside the repository and produce releaseIdentityCli prepared manifests plus per-PID start receipts for every canonical sleeve at the next launch; obtain the operator ruling on the K1-stopped sol-migration-sleeve (resume vs accept a 2-of-3 desk); then re-run the audit script embedded verbatim in the source receipt (scriptSource; sha256 recorded) to author a fresh candidate daily (24h TTL) and hand the operator the --write import
REFUSE venue-quote-fill-latency-comparison · owner EXECUTION · TTL expired at receipt time
bounded venue-vs-model audit g5vqfl_9d6176de0c6c734fc288677ac04da61f: REFUSE — the check demands quotes/fills/latency/failures COMPARED with executable venue observations for the shadow book and 0/4…
WHAT'S NEEDED: re-arm held-arm venue quoting (FILLQUOTE_TARGET=held / --held, budget-gated — operator spend sign-off) so the booked population gets fresh executable observations; add quote RTT capture to fill-quote rows and decision->quote->booked stamps to the SOL sleeves to populate latencyMs/venueLatencyMs; build the daily comparison fold (modeled fill vs matched-size venue quote + venue failure-rate vs a model-side failure artifact); then re-run this audit and re-import
REFUSE restart-outage-restore-receipts · owner OPERATIONS · decays in 64h
restart/outage/restore audit g5rorr_455097ba32449f232475078297e5f4fc executed 2026-08-29T06:22:01.394Z on this box: REFUSE — 'restart/outage/restore receipts clean' fails on receipt coverage, drill,…
WHAT'S NEEDED: let the A20 relaunch receipts accumulate a full 7-day window (they began 2026-08-28); with operator approval (standing recycle authority covers TRUE wedges only, not drills), run a scripted controlled restore drill — snapshot book state, stop one sleeve, let the watchdog relaunch it, assert arm1PositionReplay reconciliation (positions/peaks/last-prices) and ledger frontier advancement — and emit its content-addressed drill receipt bound to launch identity once COLDREAD_RELEASE_EVIDENCE_ROOT wiring and releaseIdentityCli per-PID start receipts exist; then re-run the audit script embedded verbatim in the source receipt (scriptSource; sha256 recorded) to author a fresh candidate weekly (7d TTL) and hand the operator the --write import
G6Minimal-capital pilot0/4 → flat vs priorREFUSE
REFUSE least-privilege-signer-executor · owner OPERATOR · TTL expired at receipt time
Executed signer/executor audit at HEAD df04086556eb (31 dirty files): REFUSE — no least-privilege signer or executor has been provisioned because no pilot has been authorized, BUT the surface is NOT…
WHAT'S NEEDED: before any pilot: obtain from the venue (bankr.bot) a written key-capability scope for the provisioned API key (trade-only, no-withdrawal) or re-issue a trade-only key; freeze it as an operator-signed pilot-authorization-receipt naming the exact executor path (src/index.ts Bankr NL prompts), the gates (DRY_RUN, --confirm-live), and the caps; then re-run this audit (script embedded verbatim in the source receipt, sha256 recorded) and re-author
REFUSE withdrawal-prohibition-and-limits · owner RISK · TTL expired at receipt time
Executed withdrawal-surface audit at HEAD df04086556eb: REFUSE — no withdrawal prohibition exists as a policy artifact, and the custody model makes documented absence insufficient. MEASURED: capital…
WHAT'S NEEDED: author the pilot policy BEFORE any pilot: venue-side withdrawal prohibition (trade-only key scope or venue withdrawal lock, receipt-backed from bankr.bot), plus the local limit table (MAX_TRADE_USD, DAILY_MAX_SPEND_USD, allocation caps) frozen into an operator-signed pilot-policy-receipt; add a repo-side guard that refuses any agentPrompt not matching the authorized Buy/Sell template; then re-run this audit (script embedded in the source receipt) and re-author
REFUSE human-pilot-authorization · owner OPERATOR · TTL expired at receipt time
Executed pilot-authorization census at HEAD df04086556eb: REFUSE — no human pilot authorization is on record anywhere on this desk. MEASURED: recursive name scan of data/ (depth 3) for pilot|authoriz…
WHAT'S NEEDED: none until the operator chooses to authorize a pilot; when that happens, the authorization must be a signed operator receipt naming scope (strategy, venue, capital ceiling, duration, halt criteria) frozen under data/receipts/, and this check re-authored from a fresh run of the embedded audit script against that receipt
REFUSE automatic-pilot-halt · owner RISK · TTL expired at receipt time
Executed automatic-pilot-halt audit at HEAD df04086556eb on this box: REFUSE — no pilot exists to halt, and the halt machinery that DOES exist is operator-in-the-loop, not automatic, so a…
WHAT'S NEEDED: before any pilot: build the automatic halt — a risk verdict (RK1-FIRED or a pilot loss bar) that itself writes the stop marker the pilot executor honors mid-run (not just pre-run), plus a drill seam so a synthetic trigger can traverse verdict->halt under injection; obtain the operator ruling G4 already flagged (is pager-plus-operator an accepted kill switch, or must FIRED halt); then run the injection against the DEPLOYED pilot runtime with launch-bound identity receipts and author a real failure-injection-receipt from that run
G7Scaling0/4 → flat vs priorREFUSE
REFUSE multiple-stable-forward-cohorts · owner RESEARCH · decays in 63h
forward-cohort-receipt audit g7msfc_0aab1f6076df30b5c3e2ea9b9b1cea85 over the live sleeve ledgers, stop/hold markers, the OI-110 successor funnel, and the sol-benchmark market null: REFUSE — the check…
WHAT'S NEEDED: start the stability clock properly: (a) operator resolves the K1-stopped migration arm (re-register with a fresh post-fix K1 baseline, or retire it) and rules on established-v1's hold; (b) build the pre-registered cohort-receipt fold CLI (cohort definitions, stability statistics, sol-benchmark null comparison, weekly re-authoring) so that when two-plus strategies run forward the receipts accrue automatically; (c) re-author this check weekly from that instrument — green needs multiple cohorts each stable across weeks of forward operation, which no build can shorten
REFUSE measured-capacity-and-impact · owner EXECUTION · decays in 64h
Executed capacity/impact audit g7mci_d344517f297d6c6098093ce9e669c4f3 at HEAD 91ce0f9f4096: bounded stability-frozen folds of all four quote/impact tapes plus a golden re-execution of the production…
WHAT'S NEEDED: to turn this green, in order: (a) re-authorize the size-ladder lane budget on the Titan seam (the $5/mo cap was priced for Metis; Titan quote requests measured $0/M on the real bill — re-cost bandwidth, then a priced button to the operator) and re-run sizeLadderQuoteCli over BOTH candidate and held populations, deriving impact-vs-size from executable-out retention curves since Titan carries no priceImpactPct field; (b) build the capacity fold that turns rung curves into max-deployable-size-at-tolerated-impact per strategy and compares it against arm1Fills' modeled impact at actually-traded sizes; (c) put the ladder on a cadence so this check's 7d TTL is refreshable; then re-run the audit script embedded verbatim in the source receipt (scriptSource, sha256 recorded) from the repo root on this machine and hand the operator the --write import.
REFUSE no-unresolved-p0-p1 · owner ARCHITECTURE · decays in 24h
canonical committed board still contains 16 unresolved P0/P1 item(s)
WHAT'S NEEDED: resolve every canonical P0/P1 and collect from a clean committed HEAD
UNKNOWN independent-scaling-review · owner OPERATOR · TTL expired at receipt time
no evidence imported yet — produce a fresh identity-bound review-receipt
Binding constraint: bounded-tail-clock-discipline-audit g1seic_a9d84fe4033125f466f094d5 over 7 canonical event tapes (39324 rows in 2097152-byte tail windows, newest end, current generations only; 0 parse failures): FAIL - the clock discipline is NOT desk-wide: wide-runners.log.jsonl is FROZEN_SOURCE_CLOCK. Failing: wide-runners.log.jsonl=FROZEN_SOURCE_CLOCK (3920 rows carry ONE source-clock value (2026-08-14T04:06:12.561Z) while their ingest clock spans 2434 min: coverage without validity). Passing: sol-tape.pump.jsonl=BOTH_WALL_CLOCKS, sol-tape.pumpswap.jsonl=BOTH_WALL_CLOCKS, sol-tape.pump-curve.jsonl=SOURCE_ORDINAL_ONLY, sol-tape.pump-tx.jsonl=SOURCE_ORDINAL_ONLY, pump-creates.log.jsonl=SOURCE_ORDINAL_ONLY, discovery.log.jsonl=BOTH_WALL_CLOCKS. Not audited: 6 entries listed on the source receipt.
REFUSE source-event-ingest-clocks · owner DATA · next fix the writers of the failing stream(s) (wide-runners.log.jsonl) so every row carries a valid source/event clock, then re-run --write --candidate and import · due 2026-09-03 07:16:41.025Z · stop when every canonical event tape classifies BOTH_WALL_CLOCKS or SOURCE_ORDINAL_ONLY over its audited window
3/5next-gate verified checks
0/0weekly decision conversion
0overdue unruled tranches
REFUSEdeployed identity evidence

Experiment conversion

sol-migration-early-traction-30s-v2 STALE COLLECTING
4,941observed
3,517eligible
0evaluated
0decision-grade
0decided
Excluded: MINT_MISSING 1,424
Non-evaluable action cells (not source exclusions): UNKNOWN:CONTROL_NOT_EVALUABLE; control=UNKNOWN/ENTRY_EVIDENCE_UNKNOWN/TERMINAL; challenger=EVALUABLE/NONE 2,319 · UNKNOWN:CONTROL_NOT_EVALUABLE; control=UNKNOWN/ENTRY_EVIDENCE_UNKNOWN/TERMINAL; challenger=UNKNOWN/ENTRY_EVIDENCE_UNKNOWN/TERMINAL 313 · UNKNOWN:CONTROL_NOT_EVALUABLE; control=UNKNOWN/OUTCOME_RECEIPT_MISSING; challenger=EVALUABLE/NONE 201 · UNKNOWN:CONTROL_NOT_EVALUABLE; control=UNKNOWN/OUTCOME_RECEIPT_MISSING; challenger=UNKNOWN/ENTRY_EVIDENCE_UNKNOWN/TERMINAL 29 · UNKNOWN:CONTROL_NOT_EVALUABLE; control=UNKNOWN/OUTCOME_RECEIPT_MISSING; challenger=UNKNOWN/OUTCOME_RECEIPT_MISSING 1
contractsolmasc5_572f98a58ee21abb24381af59bf82db77bedf5a29cabbc9800cb3705d15b53a7
collector30m cadence · last 2026-08-30 13:17:01.816Z · next 2026-08-30 13:47:01.816Z
maturity / owner2026-09-08 14:30:00Z · OI-110
Stop: issue exactly one governed ruling by 2026-09-08T14:30:00.000Z

Calendar trajectory

4 earlier receipt(s) use another frozen definition and are excluded rather than rescored.
08-27committed G—
08-28GAP
08-29committed G0 · dirty G—
08-30committed G0
08-31committed G—
09-01committed G—
09-02committed G0 · PARTIAL
Diagnostics are not readiness
Delivery composition, code gravity, tests, lint, process-map debt, capabilities, backups, release identity, and Structural Footprint Index v0 remain diagnostics. They cannot compensate for a failed earlier gate.
{
  "profile": "FULL",
  "canonicalDataRoot": "C:\\Users\\Jeremy\\Cwod\\bankr-robinhood-agent\\data",
  "structuralFootprintIndex": {
    "name": "Structural Footprint Index v0",
    "authority": "DIAGNOSTIC_ONLY_NOT_READINESS"
  },
  "processMap": {
    "exitCode": 10,
    "durationMs": 1831,
    "outputDigest": "7e6a35581d5d6648301f803e9c1cb0162913eabf62a883b3f42e9c571170cc4b"
  },
  "capabilityManifest": {
    "exitCode": 0,
    "durationMs": 993,
    "outputDigest": "4ae328e904caee23790f81a7bdd3f47d6dd68144a5785bb9e361e855c465852d"
  },
  "lint": {
    "exitCode": 1,
    "durationMs": 44557,
    "outputDigest": "702d4a2eca16e3fada26d6c2b30e8ecacf27fca25649635685e54d7f13d588de"
  },
  "gateDefinitionDigest": "0c30639c9cda913ede955d30376a8069c8d5e0a9a50fcd8a747cf8ec0af8908c",
  "externalEvidence": {
    "status": "VERIFIED",
    "sourcePath": "C:\\Users\\Jeremy\\Cwod\\bankr-robinhood-agent\\data\\receipts\\readiness-external-evidence.v1.jsonl",
    "physicalGeneration": "1180692972:9570149209213956:1788023854400.5332",
    "completeEof": 121159,
    "digest": "4d2716b7d42ec0430fbe300c24c7333c4f40f5a46a878db3c1eb24320b8c1db2",
    "rowCount": 31,
    "detail": "strictly validated 31 content-addressed row(s)"
  }
}

Governance clocks

ALL CLOCKS OK watch exit-state verdict · as-of 2s ago (2026-09-02 20:26Z)
stateclockt-minusboundary (utc)collector product
CLOSED oi110-four-action-holdout-open T+3.7d 2026-08-30 04:03Z v2 selector report: age 10m (bound 2.0h) | tight-trail ledger: age 77m (bound 6.0h)
CLOSED gt-paid-bulk-clamp-end T+44.4h 2026-09-01 00:00Z CG credit journal: age 0m (bound 2.0h)
CLOSED v5-migration-holdout-open T+29.9h 2026-09-01 14:30Z V5 solmaspr1 pass receipts: age 16m (bound 6.0h)
CLOSED established-held-action-dev-end T+6.6h 2026-09-02 13:50Z established-held ledger: OVERDUE age 2.5d (bound 6.0h), no advance since previous run
CLOSED exit-policy-v3-holdout-open T+5.4h 2026-09-02 15:00Z composite-exit materializer v2 receipts: OVERDUE age 2.5d (bound 6.0h), no advance since previous run
OK oi110-four-action-frozen-end T-3.3d 2026-09-06 04:03Z v2 selector report: age 10m (bound 2.0h) | tight-trail ledger: age 77m (bound 6.0h)
OK v5-collection-end T-5.8d 2026-09-08 14:30Z V5 solmaspr1 pass receipts: age 16m (bound 6.0h)
OK oi110-moratorium-end T-6.1d 2026-09-09 00:00Z v2 selector report: age 10m (bound 2.0h)
3 · EXPERIMENTS & LEARNING

Experiment scoreboard

READ-ONLY
DESCRIPTIVE ONLY Forward experiment evidencenot promotion evidence; no paper-book authority

Thesis → action learning

PROPOSAL_READY
thesis-action-learning-20260822-v2 · 2026-09-02 20:16Z · 1:16 PM
Frozen four-action selectorCOLLECTING HOLDOUT · thesis-four-action-state-conditioned-selection-v2Frozen condition SUPPORTED → INVALIDATED under decision-time policy trail-25-30-v2. 45 eligible exact outcomes / 45 global independent pool+mint clusters. Leading single proposal-eligible cell 9f7511c7:falsifier:quoteReserveVsEntry:invalidated-full-exit-next-same-pool-print-v1 runway: development 26/20, holdout 0/20, total 26/40. 0 populated mixed-path cells, 106 state/policy mismatches, and 0 outcomes missing decision-time incumbent-policy evidence are excluded from that runway · 0 alternatives excluded because the incumbent execution decision came first · 0 passing cells / 0 versioned conditional drafts across 27 frozen theory-reason-action cells. Mixed prior-transition cells: 0; mixed incumbent-policy cells: 0. Development 2026-08-23 04:03Z · 9:03 PM; holdout 2026-08-30 04:03Z · 9:03 PM; frozen end 2026-09-06 04:03Z · 9:03 PM. No cell can propose before that end; afterward every cell deterministically proposes or retains the incumbent. Every draft is NOT_ACTIVATED and cannot book, promote, mutate, or deploy a strategy.
Action-evidence conversion diagnosticsFIRST BLOCK: HOLDOUT_CLUSTER_FLOOR9,237 eligible state transitions · 2,185 transitions with a preregistered action · 395 first-invalidation opportunities · 316 full-exit actions priced · 311 full-exit outcomes complete · 415 50% reduction decisions · 304 50% reduction outcomes · 346 tight-trail decisions · 156 complete four-action outcomes · 91 condition-matched post-cutover invalidations · 63 condition-matched full-exit actions priced · 77 condition-matched tight-trail decisions · 45 condition-matched four-action outcomes · 151 conditioned post-cutover outcomes · 45 conditioned selector-eligible outcomes · 26 leading exact-cell independent clusters. Named pending/UNKNOWN/exclusion buckets: 21 state transitions outside the frozen protocol · 7,052 transitions with no preregistered alternative · 1,790 repeat invalidations represented by an earlier action · 0 registered alternatives missing opportunity evidence · 0 full-exit non-causal clock exclusions · 5 partial-reduction outcomes pending · 103 partial-reduction terminal UNKNOWN · 3 partial-reduction non-causal clock exclusions · 3 tight-trail outcomes pending · 249 tight-trail terminal UNKNOWN · 1 tight-trail non-causal clock exclusions · 0 selector outcomes outside registered cells · 0 selector outcomes missing a complete four-action set · 0 selector outcomes missing full-exit stress · 0 selector non-causal clock exclusions · 106 selector state/policy mismatches · 0 selector outcomes missing decision-time policy. These counts span separately activated zero-history boundaries; they diagnose the first blocked product stage and must not be subtracted as one historical cohort waterfall.
9,356state transitions observed
9,335inside frozen protocol
9,237causally actionable transitions
98post-exit transition refusals
47former-clock outcomes excluded
1,013independent clusters
9,149exact transition terminals
88transition outcomes pending
21named exclusions
303complete action comparisons
5action outcomes pending
9,237frozen transition decisions
0transition decisions missing
7,052transitions with no preregistered alternative
1,790repeat invalidations already represented by first action
0registered alternatives missing opportunity evidence
316registered alternatives priced
311alternative outcomes complete
Execution-decision clock integrity0 full-exit · 3 partial-reduction · 1 tight-trail exclusionsA challenger print must exist strictly before the incumbent execution decision. A later ledger booking never restores causality; excluded rows remain visible here but cannot enter descriptive cells or selection.
Preregistered action stressAVAILABLE · 297/297 post-cutover outcomes coveredPre-effective outcomes are never backfilled. Complete stress coverage and positive stressed development and holdout means are required before operator review.
Invalidation corroborationAVAILABLE · 415/415 exact forward opportunities observed · 0 missing167 confirmed · 244 contradicted · 4 unknown · 12 pre-cutover opportunities excluded. The denominator begins at the physical zero-history byte boundary. Evidence availability is clocked, so information arriving after the action print or exit cannot masquerade as decision-time knowledge.
Three-action extensionAVAILABLE · continue vs full exit vs 50% reductionForward, own-size, non-booking, descriptive only, and no retroactive selector authority. The reconciled receipt equation below separates raw observations from selector-eligible evidence.
Three-action reconciled denominator415 opportunities = 307 raw outcomes + 103 terminal UNKNOWN + 5 true pending415 opportunity rows/398 independent opportunity clusters · 415 decisions/398 independent decision clusters · raw outcomes 307/297 independent = eligible 304/294 + clock-excluded 3/3 · 3 observed ineligible · terminal UNKNOWN before/after decision 0/103 · 0 superseded terminal rows. Independent category counts are not summed when one pool+mint appears in more than one category.
Three-action terminal UNKNOWN reasonsUNKNOWN_INCUMBENT_EXIT_PRECEDED_ACTION_PRINT 34 rows/34 independent · UNKNOWN_SOURCE_GENERATION_GAP 39 rows/38 independentReason counts are derived live from the effective validated terminal set; superseded rows are excluded; UNKNOWN_SOURCE_RANGE_CEILING rows are listed separately below as a scan BOUND.
Three-action BOUND — no fill within first 64 MiB (scan cap)BOUND 30 rows/30 independentDisplay relabel only: these rows remain counted inside the 103 terminal UNKNOWN total above and their receipts are untouched. A BOUND is a statement about our scan budget, not about the market.
Three-action terminal UNKNOWN A/B/C auditA/B/C snapshot unavailable because the terminal set or frozen source identity changedNo stale audit counts are carried forward. Run a new bounded, digest-bound audit before classifying the changed terminal set.
Most-populated three-action cellinvalidated-reduce-50pct-next-print-residual-incumbent-v1 · falsifier:tape.prints.last300sn=58, independent=58 · partial vs continue +0.0049 SOL · partial vs full exit -0.0043 SOL · stressed partial vs continue -0.0031 SOL. Descriptive until a future-effective selection contract is preregistered.
Four-action tight-trail extensionAVAILABLE · 192 triggersContinue vs full exit vs 50% reduction vs full exit at the first complete executable 10% giveback from the decision-time peak. Exact own-size V2, 5%/5% stressed, non-booking, non-promotable, and descriptive only.
Four-action tight-trail reconciled denominator409 opportunities = 157 raw outcomes + 249 terminal UNKNOWN + 3 true pending409 opportunity rows/393 independent opportunity clusters · 346 decisions/332 independent decision clusters · raw outcomes 157/154 independent = eligible 156/153 + clock-excluded 1/1 · 1 observed ineligible · terminal UNKNOWN before/after decision 63/186 · 2 superseded terminal rows. Independent category counts are not summed when one pool+mint appears in more than one category.
Four-action tight-trail terminal UNKNOWN reasonsUNKNOWN_DECISION_PEAK_STATE_UNAVAILABLE 1 rows/1 independent · UNKNOWN_INCUMBENT_DECISION_PRECEDED_CHALLENGER 9 rows/9 independent · UNKNOWN_PARENT_ACTION_TERMINAL 102 rows/100 independent · UNKNOWN_PEAK_SOURCE_GENERATION_GAP 127 rows/126 independent · UNKNOWN_TERMINAL_PEAK_PRECEDES_INCUMBENT_DECISION 10 rows/10 independentReason counts are derived live from the effective validated terminal set; superseded rows are excluded.
Four-action tight-trail terminal UNKNOWN A/B/C auditA/B/C snapshot unavailable because the terminal set or frozen source identity changedNo stale audit counts are carried forward. Run a new bounded, digest-bound audit before classifying the changed terminal set.
Most-populated four-action cellinvalidated-tighten-trail-10pct-complete-own-size-peak-v1 · falsifier:quoteReserveVsEntryn=17, independent=17, triggered=12 · tight vs continue -0.0716 SOL · tight vs full exit +0.0166 SOL · tight vs partial -0.0276 SOL · stressed tight vs continue -0.0720 SOL. No selector authority.
Most-populated frozen action cellinvalidated-full-exit-next-same-pool-print-v1 · falsifier:tape.prints.last300sdevelopment n=115, nominal mean +0.0165 SOL; holdout n=63. Stress AVAILABLE · 178/178 covered · development stressed mean +0.0061 SOL · holdout stressed mean +0.0054 SOL. Descriptive until every preregistered gate passes.
Emerging reason-conditioned action splitexit-now ahead for falsifier:tape.prints.last300s: +0.0465 SOL · continue ahead for falsifier:quoteReserveVsEntry: +0.10 SOLChallenger minus incumbent, development only. Positive cell n=40, independent=40, stressed +0.0298 SOL, COLLECT. Negative cell n=78, independent=78, stressed -0.10 SOL, COLLECT. This is the live mechanical answer to why one invalidation reason can imply a different action than another, but it remains descriptive until chronological holdout and every frozen gate pass.
Bounded sparse action-context search26 qualified / 870 tested cellsDevelopment only, using the fixed one/two-field family over existing context buckets: 105 meet the independent-cluster floor, 527 retain the same sign in early and late chronological halves, 125 survive global BH-FDR, and 214 meet the directional outlier limit. Qualification can nominate only a new future experiment after the frozen collection ends; it cannot alter this experiment or trading.
Excluded transitions remain counted and reason-labelled; they cannot widen the registered theory family. No cell can book, promote, or mutate a strategy.

Cross-producer normalized admission

COLLECTING
fixed 0.05 SOL research input · source-qualified enter-all · trail-25/30 next-print outcomes · explicit zero only after a settled evaluable opportunity · 2026-08-29 00:06Z · 5:06 PM
1,987full admitted denominator
1,987classified exactly once
130complete normalized pairs
1,857unresolved, still explicit
0terminal incomplete
0invalid rows
Lifecycle denominator56 source-action UNKNOWN · 1,272 awaiting entry · 529 awaiting trail · 0 awaiting next fill · 0 lineage pending0 duplicate keys · 0 integrity issues. Missing outcomes remain pending/UNKNOWN/incomplete and are never rewritten to zero.
Forward direct identity evidence1,494 unique bounded-window receipts: 1,345 PROVEN · 149 REFUSEsol-direct-source-identity-proofs.jsonl: 1,350 canonical unique · sol-runner-direct-selection-refusals.jsonl: 144 canonical unique. Reasons: SOURCE_IDENTITY_NO_FORWARD_JOIN 144 · SOURCE_IDENTITY_PROVEN 1,345 · SOURCE_IDENTITY_QUOTE_ASSET_UNSUPPORTED 5. Producers: sol-established-dip-buy-v2 1,303 receipts: 1,303 PROVEN / 0 REFUSE · sol-runner-entry-v2 191 receipts: 42 PROVEN / 149 REFUSE. Latest receipt clock 2026-09-02 20:26Z · 1:26 PM. Runner direct-event evidence is accepted only with its current observed-sequence/frontier fence. Both ledgers are diagnostic-only companions: they grant no strategy, risk, selector, promotion, paper-book, runtime, deployment, recycle, or live-wallet authority and never rewrite a frozen source/action UNKNOWN. The normalized historical UNKNOWN count remains immutable per source/action; a later receipt never repairs it.
sol-migration-entry-v274/689 complete · 615 unresolved · 0 terminal incomplete74 independent development clusters · 2/3 positive stressed folds · triggered-pair stressed mean +6.73% (descriptive, path-conditioned) · 0 holdout clusters · not eligible. Version sol-migration-entry-v2.
sol-established-dip-buy-v246/1,175 complete · 1,129 unresolved · 0 terminal incomplete1 independent development clusters · 0/3 positive stressed folds · triggered-pair stressed mean -15.79% (descriptive, path-conditioned) · 0 holdout clusters · not eligible. Version sol-established-dip-buy-v2-thesis-exit-20260820.
sol-runner-entry-v210/123 complete · 113 unresolved · 0 terminal incomplete5 independent development clusters · 1/3 positive stressed folds · triggered-pair stressed mean -5.52% (descriptive, path-conditioned) · 0 holdout clusters · not eligible. Version sol-runner-entry-v2.
NO CANDIDATE The validated report has no comparison, selector, paper-book, strategy, risk, runtime, deployment, recycle, or live-wallet authority. PROPOSED means human review only; it never auto-activates.

Established containment

HOLD_NEW_ENTRIES
current evidence era · sol-established-dip-buy-v2-thesis-exit-20260820
1,034strict shadow decisions
802pool + mint clusters
465exact terminals
4pending outcomes
333explicit UNKNOWN
465matured clusters
independent clusters465 / 20
terminal receipts (descriptive)465 / 30
RETAIN_HOLD_NEGATIVE_EVIDENCE Hold release is not authorized. Stress, holdout, independence, and statistical gates still apply.
First decision per independent clustermean -23.0% · median -50.3% · stressed mean -29.7% · stressed median -54.6%An UNKNOWN or pending first decision is never replaced by a later exact outcome; duplicate same-pool outcomes are never summed.

Migration challenger

SHADOW
UNKNOWNpending
156evaluable
192UNKNOWN denominator
source identity: 29 forward no-mint attempts: resolved-no-coin-mint 23 · rpc-failed 6
Exact t=0 resolver outcomes only; latest source availability 2026-08-28 20:07Z · 1:07 PM. This explains future MINT_MISSING observations but cannot recover or rewrite a frozen control-time identity.

Composite exit

SHADOW
UNKNOWNhistorical authorized
UNKNOWNcumulative paired decisions
Current identity reconciliation is separate checkpoint evidence; UNKNOWN is never subtracted from this historical projection. holdout open for 5h 26m · 2026-09-02 15:00Z · 8:00 AM

Forward orchestrator

REFUSE
2026-09-02 20:10Z · 1:10 PM · failed at migration-post-feature
last supervised run; PASS means the bounded steps completed, not that a policy won
bounded read · counts cover the labelled window, not the whole history · sol-migration-early-traction-shadow.jsonl tail window: last 2.0 MiB / 348 rows of 80.4 MiB | receipts\sol-direct-source-identity-proofs.jsonl tail window: last 2.0 MiB / 1,350 rows of 52.0 MiB | sol-forward-experiment-orchestrator-receipts.jsonl tail window: last 1.0 MiB / 733 rows of 127.9 MiB | sol-composite-exit-materializer-v2-receipts.jsonl tail window: last 8.0 MiB / 1,609 rows of 19.3 MiB

Oracle & exit diagnosis
This panel folds the newest 8 MiB of the autopsy tape (OI-166 preview bound) — a WINDOW, not the whole book. Reconciled 2026-08-29 against all phase closed trades: 99.91% carry an autopsy row, 90.1% carry full judgments, 9.8% are structurally unjudgeable (no captured tape in their window). Full ledger-vs-judged reconciliation: data/proposals/ORACLE-COVERAGE-RECONCILIATION-2026-08-29.md.

biggest known leaks

1 · THE EXIT LEAK — 30 hard stops that should not have been reachable, -4.59 SOL. These trips reached a tape peak of at least 25% (median 77%), so the trail was ARMED and sat at 0.7x of peak — ABOVE the 0.5x-of-entry hard stop that nonetheless fired. For contrast the 388 hard stops that never armed cost -45.80 SOL — that is the rule working. This is the size of an execution gap, not a policy result, and the cause is undecided between engine downtime, peak-reset-on-restart, and rule ordering across a price gap. Folded live from 1070 trips in arm1-trade-autopsy.jsonl, 52m old · 16 hard-stop trips had no readable peak and are in neither bucket. Deeper: npx tsx src/cli/arm1ExitAnomalyCli.ts.
2 · THE RE-ENTRY LEAK — 61% of exits reclaimed +20% within 2h; cut-all-and-re-enter beats every partial. On the same 102 trips and the same post-exit paths: full-exit + re-entry +95.70 SOL (ex-top-1 +56.85) against the LIVE full-exit's +13.81 SOL (ex-top-1 −24.20); partial-75 +11.05, partial-50 +8.29, hold-all +2.77. UPPER BOUND — the re-entry arm buys at the next achievable print with no impact haircut and models no sell rule of its own, and 45 of 102 remainders never triggered the trail and are MARKED at the window's last price rather than realised. An unexited bag is a position, not a P&L. Producer src/cli/arm1ReentryStudyCli.ts — no artifact on disk, so this page cannot age these figures; they are stated as of the 2026-08-10 read and re-derive with npx tsx src/cli/arm1ReentryStudyCli.ts. An instrument with no artifact is invisible to every idle detector here, which is why it is listed above under on-demand readers.
Best REALIZABLE exit per closed trip vs what we actually got. OWN-SIZE DEPTH IS CHARGED — the oracle's argmax runs over what our actual heldBase would have received at each mark's real pool reserves (arm1Fills), so a taller print in a thinner pool can lose to a lower print in a deeper one. The superseded fee-only bound totalled +891.21 SOL; 98% of it survives the charge. A trip whose depth cannot be priced is UNMEASURED and excluded, never folded in at zero impact. CURRENT SOL PHASE 2 arm1-oracle-outcomes.jsonl
61%median regret / trip
+874.16 SOLtotal regret
+859.75 SOLex-top-1
3109trips labeled

current exit-policy forward cohort

trail-25-30-v2 is segmented by the exact exitPolicy receipt on each ledger exit, not inferred from rule name or time. Realized figures cover all 3495 derived exits; oracle figures cover only the 3037 strict tape-joined exits. This is a forward operational cohort, not a randomized comparison: positions opened before deployment can close under the new manager.
3495exits managed
-156.61 SOLrealized SOL
31%win rate
-12.0%median return
-173.58 SOLrealized ex-best
3037/3121oracle joined / folded
61%median oracle regret
+889.03 SOLoracle regret ex-top-1
Coverage is part of the result. Strict learning join: 3119 joined trips · 87 folded autopsy rows without oracle · 225 oracle rows without folded autopsy. This panel is the observed subset, not the whole ledger.

median regret by exit rule

exit rulenmedian regret
hard-stop116273%
trail173454%
operator-close2135%
time-stop19225%

attribution — entry problem or exit problem

trail-25-30-v2 joined subset — this is the exact policy cohort inside the selected current-phase scope.
ENTRY-side (never worked) · current policyn=1074 · -219.63 SOL
EXIT-side (round-tripped + clipped) · current policyn=1090 · +151.54 SOL
ENTRY-side (never worked) · current-phasen=1103 · -225.04 SOL
EXIT-side (round-tripped + clipped) · current-phasen=1115 · +151.70 SOL

worst cells — exit rule × path class · current phase

ruleclassnrealizedmed peakmed giveback
hard-stopnever-worked915-220.58 SOL-19%—
hard-stopshallow212-61.72 SOL6%1%
trailshallow496-20.01 SOL15%14%
hard-stopround-tripped57-18.72 SOL66%99%

policy ranking — CURRENT SOL PHASE 2 (1048 readable paths)

Same counterfactual method as the archived book's “rule stack”, on the book that is trading. SOL-weighted. trail-25/30 IS the live replay arm (trail-25-30-v2) — it is scored here so the deployed rule has a place in its own ranking. OWN-SIZE DEPTH IS CHARGED — every arm (and ACTUAL) is haircut at its own fill mark's real pool reserves via arm1Fills (OI-4). Arms still fill at a print while ACTUAL carries real fills, so arm-vs-arm remains the sound comparison.
policySOL% of deployedfired
ACTUAL (what we ran)-57.32 SOL-30.0%—
support-5/2-28.29 SOL-14.8%981/1048
fade-3-31.28 SOL-16.4%999/1048
support-10/5-31.60 SOL-16.5%932/1048
trail-25/10-39.21 SOL-20.5%399/1048
trail-50/10-43.76 SOL-22.9%237/1048
trail-50/15-45.31 SOL-23.7%237/1048
stop-70pct-45.32 SOL-23.7%770/1048
trail-50/20-46.05 SOL-24.1%236/1048

trade learning — CURRENT SOL PHASE 2 (1070 judged trips)

Same judge as the archived USD book (tradeVerdictCore.judgeTrade), fed this book's own tape-anchored path fold. Pre-entry/post-exit study windows: 60min each, matching the USD book's own convention.
1070trips judged
1048with a verdict
1045judgeable entries
1045judgeable exits
EXITS SKEW LATE — 292 too-late vs 66 too-early (4.4×), plus 285 bad in both directions. The book's exit failure is holding THROUGH the peak, not selling before it. Estimated replayable impact (best-policy benchmark, SOL): too-late cohort +38.53 SOL total (+0.13 SOL/trade) vs too-early cohort +0.27 SOL total (+~0.0041 SOL/trade).

entry verdict

doomed430
well-timed304
early218
late93
unknown3

exit verdict

well-timed402
too-late292
too-early66
both285
unknown3
3 entries and 3 exits are UNKNOWN — the honest "not enough evidence" state, counted so you can see how much of the book is still unjudgeable, not hidden to make the bars look cleaner.

per-trip oracle exits — worst regret first

For each closed trip: WHEN the best realizable exit was available, what it was worth, what we actually got, and the gap. "oracle at" is the instant a decision there would have been fillable at the NEXT print — you cannot sell into the print that told you to sell.
tokenour exitoracle atoracle %we got %regretheld
utility SOL htNct9u2iZxsGN3NGopxfgeTd54fRU9ZevxubPWpump trail +71m in 3943% 523% +14.41 SOL 75m
Fwog SOL CmPqbqnkpjGsmJxJHG1ZZ4tLNXSy5nGHw2a9b1vkpump trail +58m in 4084% 2261% +13.69 SOL 59m
BABYANSEM SOL 8vQ7Ee5vZXfuw8JzUGzYMsyz84QWjDxmdcFPN4b5pump hard-stop +2m in 420% -53% +4.84 SOL 2m
vibes SOL ELQHz8ju1aicZx6GPA9YAZn7eX5AVLGbuEnyt4TWpump trail +21m in 446% -43% +4.36 SOL 21m
AB SOL BcEmfDVLEVznvs5BrSr7nrxoekLbQiuNw7WdWrfQpump hard-stop +6m in 336% -54% +3.75 SOL 11m
CONDOM SOL 7Nyi5WPvWM7YgvzVhLpeQv5qLwBPJnVDUcyiKuHEpump hard-stop +3m in 339% -95% +3.41 SOL 7m
UNKNOWN SOL EyYtNeMJqTKFz1HcZYCqpFd5eqhtPYcXQ2gpAbadpump trail +7m in 2246% 62% +3.39 SOL 7m
Charlie SOL GSokbJVA1P9tAPffqpiTN7b4XnEh8mNKQQ4cBkNNpump trail +19m in 767% 406% +3.04 SOL 19m
ZION SOL 7RTVnPY8yDnoywLJmk3Lf3VtiQDwu2FCWL6HvsfLrFZ1 trail +17m in 702% 357% +2.77 SOL 38m
drooler SOL 6MAjC3uWMzGc4LAeKF3EoC85DUtX67MCFE3numnmpump trail +10m in 298% 40% +2.62 SOL 12m
Pikachu SOL CXKuSxyoY1FF42eS2BLr7gJrTK1hLrdmbho8D26Npump trail +27m in 535% 140% +2.42 SOL 28m
UNKNOWN SOL UPwxM8cK6u4WP8dzs7T3oAKgJQVQMaGSixakKtkpump hard-stop +2h 36m in 554% -98% +2.40 SOL 2h 47m
TOKEN SOL GPjTJykiPxuAqLsveAzAimFuupDYvkB4md4D12hNx5Fi hard-stop +37s in 195% -57% +2.21 SOL 44s
NIUNAI SOL H9fikySbqRei3wxMdjtW9GHTdp8r8h4SZma2dsUwpump trail +13m in 585% 306% +2.19 SOL 14m
Quantdev SOL EP5k7WwMYXPNzkAEc8wtUm2imx3sg3NDRNxyt9Ktpump trail +31m in 471% 210% +2.14 SOL 33m
Per-trip oracle rows in this render (3109 phase rows, with pool + oracle mark index; artifact may contain older rows too): data/arm1-oracle-outcomes.jsonl
Rule 39: exit-rule names are OUR OWN, so this table is a DIAGNOSTIC — never evidence that a new rule should exist. Deeper: npx tsx src/cli/arm1ExitAnomalyCli.ts · src/cli/arm1ReentryStudyCli.ts

recent autopsies — CURRENT SOL PHASE 2 (12 newest of 1070 folded)

Per-trip autopsy of the live SOL book with the 3-Grade framework (Grade 1 Execution, Grade 2 Opportunity, Grade 3 Intrinsic Potential) — path class, entry/exit timing, ceiling capture, and own-size oracle. Menu returns are DEPTH-CHARGED at each arm's own fill mark (OI-4). Percents are unit-free tape ratios; SOL figures are the ledger's own. data/arm1-trade-autopsy.jsonl
DEG SOL 7HJCxCMTnit9rLS21e9SMdisJaaZ9A5P3N1Lq99roeTz never-worked trail · held 1h 33m · 2026-09-02 18:45Z · 11:45 AM
SCORES Opp 0/100 Trade 16/100 Combined 8/100
TRADE entry doomed · exit well-timed · pre-entry runup +424%
OPPORTUNITY ENTRY left $ on table ceiling +88.3% · 16% captured · +0.17 SOL left on table
why miss: better entry alone +424pp · better exit alone +22pp
CAPTURE partial · 16% of ceiling · ceiling +88.3%
peak -64% at +3.0min → exit -86.5% · realized -0.01 SOL on 0.10 SOL
oracle -64% realizable · regret +22%
entry DOOMED: the best FILLABLE exit anywhere on this path was -64%, so no exit rule could have made money on it — 424% of the move was already spent before our fill. This is an ENTRY-SELECTION failure, not a timing one — the only fix is not taking it; exit well-timed: sold into an unwinnable path (no fillable exit cleared cost — timing is not the fault); REGRET 22 pts ($0.02): the best realizable exit on this path was -64% vs our -87% (closest menu rule fade-3 at -79%)
better menu exits if fired: fade-3 -79.3% · support-10/5 -80.1% · stop-50pct -81.6%
UNDEADS SOL EguENXBejC9EvHqqfKZ1koVnUNbbYM1EcLsPcJtxpump never-worked trail · held 3m · 2026-09-02 16:43Z · 9:43 AM
SCORES Opp 0/100 Trade 20/100 Combined 10/100
TRADE entry doomed · exit both · pre-entry runup +633% · post-exit continuation +26%
OPPORTUNITY ENTRY left $ on table ceiling +47.6% · 20% captured · +0.13 SOL left on table
why miss: better entry alone +633pp · better exit alone +57pp
CAPTURE partial · 20% of ceiling · ceiling +47.6%
peak -26% at +2.0min → exit -84.4% · realized -0.04 SOL on 0.10 SOL
oracle -27% realizable · regret +57%
entry DOOMED: the best FILLABLE exit anywhere on this path was -27%, so no exit rule could have made money on it — 633% of the move was already spent before our fill. This is an ENTRY-SELECTION failure, not a timing one — the only fix is not taking it; exited BADLY BOTH WAYS: peaked +-26%, gave back 58 points, then it ran another +26% without us; REGRET 57 pts ($0.06): the best realizable exit on this path was -27% vs our -84% (closest menu rule fade-3 at -49%)
better menu exits if fired: fade-3 -49.0% · stop-50pct -49.3% · stop-60pct -49.3%
DEG SOL 7HJCxCMTnit9rLS21e9SMdisJaaZ9A5P3N1Lq99roeTz never-worked hard-stop · held 2m · 2026-09-02 16:33Z · 9:33 AM
SCORES Opp 0/100 Trade 4/100 Combined 2/100
TRADE entry doomed · exit both · pre-entry runup +157% · post-exit continuation +48%
OPPORTUNITY ENTRY left $ on table ceiling +57.5% · 4% captured · +0.11 SOL left on table
why miss: better entry alone +157pp · better exit alone +40pp
CAPTURE partial · 4% of ceiling · ceiling +57.5%
peak -11% at +1.4min → exit -51.6% · realized -0.05 SOL on 0.10 SOL
oracle -12% realizable · regret +40%
entry DOOMED: the best FILLABLE exit anywhere on this path was -12%, so no exit rule could have made money on it — 157% of the move was already spent before our fill. This is an ENTRY-SELECTION failure, not a timing one — the only fix is not taking it; exited BADLY BOTH WAYS: peaked +-11%, gave back 41 points, then it ran another +48% without us; REGRET 40 pts ($0.04): the best realizable exit on this path was -12% vs our -52% (closest menu rule stop-60pct at -44%)
better menu exits if fired: stop-60pct -44.5% · stop-70pct -44.5% · stop-50pct -51.0%
pippo SOL Gffw364rz4r93aYum3BHynoi5iw1gsq2m4P2Py6gpump never-worked hard-stop · held 1h 41m · 2026-09-02 14:54Z · 7:54 AM
SCORES Opp 0/100 Trade 13/100 Combined 7/100
TRADE entry doomed · exit well-timed · pre-entry runup +84%
OPPORTUNITY no real move available ceiling +4.2% · 13% captured · +0.07 SOL left on table
why miss: better entry alone +84pp · better exit alone +43pp
CAPTURE partial · 13% of ceiling · ceiling +4.2%
peak -26% at +11.1min → exit -70.0% · realized -0.05 SOL on 0.10 SOL
oracle -27% realizable · regret +43%
entry DOOMED: the best FILLABLE exit anywhere on this path was -27%, so no exit rule could have made money on it. This is an ENTRY-SELECTION failure, not a timing one — the only fix is not taking it; exit well-timed: sold into an unwinnable path (no fillable exit cleared cost — timing is not the fault); REGRET 43 pts ($0.04): the best realizable exit on this path was -27% vs our -70% (closest menu rule fade-3 at -31%)
better menu exits if fired: fade-3 -31.3% · stop-70pct -31.4% · support-5/2 -37.3%
☉ SOL GyaBdGkrktJdJJcqoXi7mVC37d8eECu5SetPz8TSpump never-worked hard-stop · held 72m · 2026-09-02 11:23Z · 4:23 AM
SCORES Opp 0/100 Trade 16/100 Combined 8/100
TRADE entry doomed · exit well-timed · pre-entry runup +765% · post-exit continuation +24%
OPPORTUNITY ENTRY left $ on table ceiling +16.8% · 16% captured · +0.09 SOL left on table
why miss: better entry alone +765pp · better exit alone +60pp
CAPTURE partial · 16% of ceiling · ceiling +16.8%
peak -18% at +15.7min → exit -79.3% · realized -0.05 SOL on 0.10 SOL
oracle -19% realizable · regret +60%
entry DOOMED: the best FILLABLE exit anywhere on this path was -19%, so no exit rule could have made money on it — 765% of the move was already spent before our fill. This is an ENTRY-SELECTION failure, not a timing one — the only fix is not taking it; exit well-timed: sold into an unwinnable path (no fillable exit cleared cost — timing is not the fault); REGRET 60 pts ($0.06): the best realizable exit on this path was -19% vs our -79% (closest menu rule fade-3 at -34%)
better menu exits if fired: fade-3 -34.1% · stop-60pct -41.4% · stop-70pct -43.5%
Biu SOL 4rxNSU5BynixWPZvN3Nprvm4fdomjRSQynFb5XvQpump shallow hard-stop · held 39m · 2026-09-02 10:47Z · 3:47 AM
SCORES Opp 44/100 Trade 40/100 Combined 42/100
TRADE entry well-timed · exit too-late · pre-entry runup +15%
OPPORTUNITY EXIT left $ on table ceiling +22.8% · 40% captured · +0.12 SOL left on table
why miss: better entry alone +15pp · better exit alone +109pp
CAPTURE partial · 40% of ceiling · ceiling +22.8%
peak +10% at +19.0min → exit -99.9% · realized -0.05 SOL on 0.10 SOL
oracle +9% realizable · regret +109%
entry well-timed; exited LATE: gave up 109 fillable points — +9% was available on this path and we took -100%; REGRET 109 pts ($0.11): the best realizable exit on this path was +9% vs our -100% (closest menu rule support-5/2 at -20%)
better menu exits if fired: support-5/2 -19.7% · support-10/5 -19.7% · fade-3 -19.7%
Bros SOL 7VwjRc8it6s3J19NFvc31To7NZvfJjiHvAM2z9KTpump never-worked hard-stop · held 63m · 2026-09-02 08:20Z · 1:20 AM
SCORES Opp 0/100 Trade 12/100 Combined 6/100
TRADE entry doomed · exit well-timed · pre-entry runup +120% · post-exit continuation +15%
OPPORTUNITY no real move available ceiling +0.2% · 12% captured · +0.05 SOL left on table
why miss: better entry alone +120pp · better exit alone +45pp
CAPTURE partial · 12% of ceiling · ceiling +0.2%
peak -2% at +0.9min → exit -47.3% · realized -0.05 SOL on 0.10 SOL
oracle -2% realizable · regret +45%
entry DOOMED: the best FILLABLE exit anywhere on this path was -2%, so no exit rule could have made money on it — 120% of the move was already spent before our fill. This is an ENTRY-SELECTION failure, not a timing one — the only fix is not taking it; exit well-timed: sold into an unwinnable path (no fillable exit cleared cost — timing is not the fault); REGRET 45 pts ($0.04): the best realizable exit on this path was -2% vs our -47% (closest menu rule fade-3 at -5%)
better menu exits if fired: fade-3 -4.6% · support-5/2 -7.2% · support-10/5 -8.9%
Hashcash SOL AL63Th2HZAjHzU5hvRiqsDumfYp3CEnZUQaFSDnVotQn round-tripped trail · held 13m · 2026-09-02 06:55Z · 11:55 PM
SCORES Opp 100/100 Trade 33/100 Combined 67/100
TRADE entry late · exit both · pre-entry runup +122% · post-exit continuation +96%
OPPORTUNITY ENTRY left $ on table ceiling +175.4% · 33% captured · +0.16 SOL left on table
why miss: better entry alone +122pp · better exit alone +68pp
CAPTURE partial · 33% of ceiling · ceiling +175.4%
peak +85% at +9.8min · giveback +82% of peak → exit +15.3% · realized +0.06 SOL on 0.10 SOL
oracle +84% realizable · regret +68%
entered LATE: 122% of the move was already spent in the 60min before our fill; exited BADLY BOTH WAYS: peaked +85%, gave back 70 points, then it ran another +96% without us; REGRET 68 pts ($0.07): the best realizable exit on this path was +84% vs our 15% (closest menu rule trail-50/15 at 55%)
better menu exits if fired: trail-50/15 +54.9% · trail-50/20 +44.3% · trail-50/10 +43.7%
Pill SOL 4232dGVrUST2LsZVJweoGmSmeqBtQdRe25aEUm3jb43E never-worked trail · held 2h 24m · 2026-09-02 06:51Z · 11:51 PM
SCORES Opp 0/100 Trade 18/100 Combined 9/100
TRADE entry doomed · exit well-timed · pre-entry runup +142%
OPPORTUNITY ENTRY left $ on table ceiling +118.3% · 18% captured · +0.17 SOL left on table
why miss: better entry alone +142pp · better exit alone +37pp
CAPTURE partial · 18% of ceiling · ceiling +118.3%
peak -14% at +57.1min → exit -52.8% · realized +0.01 SOL on 0.10 SOL
oracle -15% realizable · regret +37%
entry DOOMED: the best FILLABLE exit anywhere on this path was -15%, so no exit rule could have made money on it — 142% of the move was already spent before our fill. This is an ENTRY-SELECTION failure, not a timing one — the only fix is not taking it; exit well-timed: sold into an unwinnable path (no fillable exit cleared cost — timing is not the fault); REGRET 37 pts ($0.04): the best realizable exit on this path was -15% vs our -53% (closest menu rule support-10/5 at -38%)
better menu exits if fired: support-10/5 -37.9%
Uptember SOL U29rFVTgvGpYuKRe2YYSkjzNkz9J1D5CCeUfVGopump never-worked time-stop · held 12h 0m · 2026-09-02 06:29Z · 11:29 PM
SCORES Opp 0/100 Trade 0/100 Combined 0/100
TRADE entry doomed · exit well-timed · pre-entry runup +590%
OPPORTUNITY no real move available ceiling +11.3% · 0% captured · +0.10 SOL left on table
why miss: better entry alone +590pp
CAPTURE missed · 0% of ceiling · ceiling +11.3%
peak -84% at +17.9min → exit -84.3% · realized +~0.00071 SOL on 0.10 SOL
oracle -84% realizable · regret +0%
entry DOOMED: the best FILLABLE exit anywhere on this path was -84%, so no exit rule could have made money on it — 590% of the move was already spent before our fill. This is an ENTRY-SELECTION failure, not a timing one — the only fix is not taking it; exit well-timed: sold into an unwinnable path (no fillable exit cleared cost — timing is not the fault); REGRET 0 pts ($0.00): the best realizable exit on this path was -84% vs our -84%
Hashcash SOL AL63Th2HZAjHzU5hvRiqsDumfYp3CEnZUQaFSDnVotQn never-worked hard-stop · held 5m · 2026-09-02 06:04Z · 11:04 PM
SCORES Opp 0/100 Trade 13/100 Combined 7/100
TRADE entry doomed · exit both · pre-entry runup +153% · post-exit continuation +64%
OPPORTUNITY no real move available ceiling +10.6% · 13% captured · +0.09 SOL left on table
why miss: better entry alone +153pp · better exit alone +67pp
CAPTURE partial · 13% of ceiling · ceiling +10.6%
peak -13% at +0.2min → exit -81.3% · realized -0.05 SOL on 0.10 SOL
oracle -14% realizable · regret +67%
entry DOOMED: the best FILLABLE exit anywhere on this path was -14%, so no exit rule could have made money on it — 153% of the move was already spent before our fill. This is an ENTRY-SELECTION failure, not a timing one — the only fix is not taking it; exited BADLY BOTH WAYS: peaked +-13%, gave back 68 points, then it ran another +64% without us; REGRET 67 pts ($0.07): the best realizable exit on this path was -14% vs our -81% (closest menu rule support-5/2 at -20%)
better menu exits if fired: support-5/2 -19.9% · fade-3 -20.4% · support-10/5 -31.2%
alon SOL 72NtVkCkBC2mKCT6paAdo9g4KptUrRH8aRU3xzBVQHDW never-worked hard-stop · held 3h 31m · 2026-09-02 05:59Z · 10:59 PM
SCORES Opp 0/100 Trade 29/100 Combined 15/100
TRADE entry doomed · exit well-timed · pre-entry runup +91%
OPPORTUNITY no real move available ceiling -1.7% · 29% captured · +0.08 SOL left on table
why miss: better entry alone +91pp · better exit alone +34pp
CAPTURE partial · 29% of ceiling · ceiling -1.7%
peak -51% at +5.7min → exit -85.0% · realized -0.05 SOL on 0.10 SOL
oracle -51% realizable · regret +34%
entry DOOMED: the best FILLABLE exit anywhere on this path was -51%, so no exit rule could have made money on it. This is an ENTRY-SELECTION failure, not a timing one — the only fix is not taking it; exit well-timed: sold into an unwinnable path (no fillable exit cleared cost — timing is not the fault); REGRET 34 pts ($0.03): the best realizable exit on this path was -51% vs our -85% (closest menu rule stop-50pct at -51%)
better menu exits if fired: stop-50pct -51.5% · stop-60pct -51.5% · stop-70pct -51.5%

Learning loop pulse
Silent-failure detector. Green = evidence inside SLA. Amber = stale/idle. Red = frozen. Retired = producer deliberately stopped by ruling — history, not failure, never counted red. Unknown = artifact never written. Lanes cover the trade-learning / off-book / theories / applied machinery (the archived USD boards themselves no longer render; see the ARCHIVED USD BOOK pointer card). Trade-learning rows track the LIVE SOL book (arm ledger / arm autopsy / tape capture / shadow exits).
Scope, so a red row is not read as the live book failing: trade-learning is the LIVE SOL book. off-book / theories / applied are the decider + experiment machinery, which still targets the ARCHIVED USD book's config/strategy.*.json — Arm 1's exit rules are constants and this machinery has never changed them. The live book's own analysis instruments, with their scheduler stamps, are summarized in the live SOL book, Open positions, and Oracle/exit sections.
Discipline debt (ratcheted, growth refused by the suite): 84 runnerless CLIs · 23 rule-39 queue entries · commit A/E rate 15% doctrine-drift.log.jsonl
3/4trade green
1/3off-book green
2theories ok/idle
2red stages

trade learning

Fills & closes (arm)evidence 2m ago · arm1-engine-ledger.jsonl
Live autopsy (arm)evidence 53m ago · arm1-trade-autopsy.jsonl
Held tape captureevidence 0s ago · sol-tape.pumpswap.jsonl
Shadow exits (open)last complete batch 4d ago · held 2 positions AT THAT TIME · 0 open now — a flat book makes this tool exit without writing, so quiet here is correct, not broken (batch clock, not file mtime — the engine writes buyer-exit rows into the same file) · arm1-shadow-exits.jsonl

off-book

Fillability shadowDEAD WITH PRODUCER (retired, not broken) — every write site is inside paperRunner, STOPPED by operator ruling 2026-08-10. It cannot produce again on the live book. Live successor for veto-level refusals: arm2-abstentions.jsonl (see src/entryFunnelReadCli.ts). · entry-fillability-shadow.log.jsonl
Miss memory16392 blocked · 12444 observable · 10 biggest misses · miss-memory-latest.json
Flywheel deciderevidence 17m ago · flywheel-decider-events.jsonl

theories

Live theories3 live in traffic · active-experiments.json
Theory queue53 queue rows · experiment-queue.json

applied

Autopilot / operator actionsevidence 25.6d ago · autopilot-actions.log.jsonl
Promoted knobs overlay7 overlay entries · promoted-knobs.json
Tuning proposalson decider — refresh every ~20m; bridge → theories · tuning-proposals-auto.json
Exit-10 recycleexit-10 on escalations with NO recycle action logged — change may be inert (pre-auto-recycle or recycle never ran) · flywheel-decider-events.jsonl

what we are training

All 9 live-book instruments produced evidence inside their SLA. 1 more is correctly quiet because there is no work (marked below) — that is not counted as idle.
Engine ledger LIVE 1m old
arm1-engine-ledger.jsonl · event-driven (market) · no scheduler stamp · producer src/arm1Engine · run-arm1-engine.cmd
Held tape capture LIVE 0s old
sol-tape.pumpswap.jsonl · continuous · no scheduler stamp · producer tape2 held tier
Trade autopsy LIVE 52m old
arm1-trade-autopsy.jsonl · hourly (watchdog arm chain, stage 1) · dispatcher fired 53m ago · producer src/cli/arm1TradeAutopsyCli.ts --json
Paired policy outcomes LIVE 49m old
sol-policy-outcomes.jsonl · hourly (watchdog arm chain, stage 2) · dispatcher fired 53m ago · producer src/cli/solPolicyOutcomeCli.ts
a fresh artifact may still report zero receipt/stress coverage; read POLICY SELECTION in Work queue for the gate verdict, never infer promotion from freshness
Policy selection gate LIVE 49m old
sol-strategy-selection.json · hourly (watchdog arm chain, stage 3) · dispatcher fired 53m ago · producer src/cli/solStrategySelectionCli.ts
this is a claim/promotion control, not a Phase-2 paper-sizing alarm; its exact contract and blockers appear in Work queue
Oracle outcomes LIVE 1h 31m old
arm1-oracle-outcomes.jsonl · six-hour full-history research leg · dispatcher fired 76m ago · producer src/cli/arm1OracleOutcomesCli.ts
Learning fold LIVE 1h 30m old
arm1-learning-fold.json · six-hour full-history research leg · dispatcher fired 76m ago · producer src/cli/arm1LearningFoldCli.ts
Shadow exits (open) IDLE — NO WORK 4d old
arm1-shadow-exits.jsonl · 15 min (watchdog) · dispatcher fired 13m ago · producer src/cli/liveShadowExitCli.ts
aged by the latest run-summary batch, NOT by file mtime — the engine also appends buyer-exit rows here
Quiet is correct here: the book is FLAT — this tool verdicts OPEN positions and exits without writing when nothing is held, so a still artifact is the correct output. The age above is how old the last real reading is, not a fault.
Doctrine drift LIVE 52m old
doctrine-drift.log.jsonl · hourly (watchdog) · dispatcher fired 53m ago · producer src/cli/doctrineDriftCli.ts

on-demand readers — no artifact, so this page cannot age them

These run by hand and print to stdout. Nothing on disk changes when they run, so no staleness detector on this desk — including this one — can tell you the last time anyone looked. That is a real gap, listed rather than hidden.
src/cli/arm1ExitAnomalyCli.tshard stops that fired on trips whose trail had already armed
src/cli/arm1ReentryStudyCli.tswhat the tape did AFTER each exit — the re-entry opportunity
src/cli/entryFunnelReadCli.tsdiscovery → veto → admit funnel, Arm 1 vs the archived book

Breakdowns · closed trades
1 historical unmatched exit(s) remain quarantined (OI-128, already corrected) and are excluded from this table. They do not hide the rest of the book.
closed trades in window: 5537 · win = realizedSol > 0

by close rule

rulenwinPnL
hard-stop28417% (199/2841)-538.60 SOL
trail236652% (1223/2366)+295.13 SOL
time-stop30612% (37/306)+5.17 SOL
operator-close2214% (3/22)-1.73 SOL
buyer-exit2100% (2/2)+0.81 SOL

by entry strategy

strategynwinPnL
unknown215220% (433/2152)-90.03 SOL
birth-entry175534% (597/1755)-65.04 SOL
dip-buy96031% (296/960)-57.54 SOL
runner65120% (131/651)-26.48 SOL
legacy-arm:established-v11937% (7/19)-0.13 SOL
USD-era cuts (by profile, by exit-param path) removed 2026-08-29 — Arm 1 has one profile and constant exit rules; how each exit rule and candidate policy performed lives in Oracle & exit diagnosis.
4 · OPERATIONS

Work queue · Open items
In flight & blocked — 20 in flight · 8 blocked (compact; full detail once, below)
P0 OI-110 · PLATFORM · 21d old · CURRENT STATE 2026-09-01T15:25Z, replacing a note that read as resolved. THE EVIDENCE PROGRAMME THIS ITEM…
P1 OI-172 · PLATFORM · 7d old · refreshed 2026-08-31: operator ruled four buttons 2026-08-30 19:30 MST - G4 shaping caps adopted, drawdown…
P1 OI-178 · PLATFORM · 3d old · NEXT: lane-executable residuals remain after the rotation fix - unread archives and the thirteen unaudited…
P1 OI-180 · PLATFORM · 3d old · NEXT FALSIFICATION: spawn the real CLI post-fix with exact STOPPED markers, no receipts and no cursor and…
P1 OI-184 · REVIEW · 3d old · REVIEW 2026-09-02T19:00Z: fuller ask designed on the board. Under the lock the engine appends the book row,…
P1 OI-182 · REVIEW · 3d old · REVIEW 2026-09-02T09:30Z MEASURED: the ba04d7ff seam is active (both engines restarted after it). Runner…
39 open item(s), live from data/open-items.json. This is a presentation of the canonical board, not a second backlog. In flight is the active total; implementation and research stay separate.
20in flight
8implementation
12research
3validation
8blocked
8queued
0unclassified
0ready to close
5recently closed
NEXT ACTIVE OI-110 — CURRENT STATE 2026-09-01T15:25Z, replacing a note that read as resolved. THE EVIDENCE PROGRAMME THIS ITEM DEPENDS ON IS FAILING ON TWO FRONTS. Epoch thesis-action-v2-same-row-20260815 is VOID-IMMINENT: 423 development clusters collected but the leading proposal-eligible cell holds 1 of 20, because the protocol admits only one exact prior state per cell and 420 of 423 clusters sit in mixed cells - the only cells satisfying the purity rule are the three whose single state is UNKNOWN, so it is satisfiable mainly where observation FAILED. Epoch 2 (state-conditioned v2) does NOT have that defect, conditions on SUPPORTED up front, zero mixed exclusions - and still misses its floor by arithmetic: leading cell 26/20 development but 0/20 holdout, 9 holdout clusters at 3.80/day across all cells, and even if EVERY remaining cluster landed in that one cell it reaches 17.6 against 20 with the horizon frozen 2026-09-06. PREREQUISITE ON RELEASING THE ENTRY HOLD: the established engine predates its own liveness receipt (process started 2026-08-31T11:09Z, receipt code landed 17:15Z), so exit-invariant cannot vouch for that arm; harmless now at zero open positions, not harmless once it holds any. NEEDS AN OPERATOR RULING on the unit of account - accept that these epochs cannot conclude, or preregister a successor sized against the OBSERVED 3.80 clusters/day rather than a floor chosen before the rate was known.
next queued: OI-170 — 2026-09-01T23:45Z: operator chose in chat (~4:33 PM MST, button 'Snapshot into git') to commit CODEX's unfinished third repair whole; landed as a75c9a86 (10 files, gate green) on top of 7a6c1ed4 (the test slice). HEAD is one coherent state again. Next for PLATFORM: read the third repair from a75c9a86 against CODEX's last DM (21:38Z: emitted-JS truth gap, ambient declare erasure, fetch/timer/socket/channel/scheduler globals), run the hostile capability gate, and either finish the repair or revert both OI-170 commits' hunks to c9611381a. Unreviewed until then; census REFUSE, activation false by its own contract.

OUTCOME WORKSTREAMS 2

one canonical queue, grouped by the result each OI advances
Strategy evaluation 1 open — falsifiable challengers, selection, and strategy-specific evidence
OI-110 (P0, ACTIVE_IMPLEMENTATION)
Unassigned 38 open — legacy OI awaiting explicit portfolio classification
OI-170 (P1, QUEUED) · OI-172 (P1, ACTIVE_IMPLEMENTATION) · OI-177 (P1, BLOCKED_DEPENDENCY) · OI-178 (P1, ACTIVE_IMPLEMENTATION) · OI-180 (P1, ACTIVE_IMPLEMENTATION) · OI-181 (P1, BLOCKED_DEPENDENCY) · OI-182 (P1, ACTIVE_RESEARCH) · OI-183 (P1, REVIEW_COMPLETE) · OI-184 (P1, ACTIVE_IMPLEMENTATION) · OI-192 (P1, QUEUED) · OI-195 (P1, BLOCKED_DEPENDENCY) · OI-196 (P1, REVIEW_COMPLETE) · OI-198 (P1, ACTIVE_IMPLEMENTATION) · OI-210 (P1, QUEUED) · OI-219 (P1, ACTIVE_RESEARCH) · OI-171 (P2, ACTIVE_RESEARCH) · OI-185 (P2, ACTIVE_RESEARCH) · OI-193 (P2, QUEUED) · OI-204 (P2, ACTIVE_IMPLEMENTATION) · OI-205 (P2, BLOCKED_DEPENDENCY) · OI-206 (P2, BLOCKED_DEPENDENCY) · OI-215 (P2, QUEUED) · OI-217 (P2, ACTIVE_RESEARCH) · OI-220 (P2, ACTIVE_RESEARCH) · OI-173 (P3, ACTIVE_RESEARCH) · OI-174 (P3, ACTIVE_RESEARCH) · OI-186 (P3, BLOCKED_DEPENDENCY) · OI-187 (P3, QUEUED) · OI-189 (P3, ACTIVE_IMPLEMENTATION) · OI-190 (P3, ACTIVE_RESEARCH) · OI-191 (P3, ACTIVE_RESEARCH) · OI-194 (P3, ACTIVE_RESEARCH) · OI-197 (P3, REVIEW_COMPLETE) · OI-201 (P3, ACTIVE_RESEARCH) · OI-202 (P3, BLOCKED_DEPENDENCY) · OI-203 (P3, BLOCKED_DEPENDENCY) · OI-214 (P3, QUEUED) · OI-221 (P3, QUEUED)

IMPLEMENTATION 8

code, integration, or operating work under way
NEXT P0 CURRENT OWNER: PLATFORM OI-110 · 21d old
ESTABLISHED-STRATEGY REWORK AND PAPER-FUNDING ADEQUACY: determine why established-v1 loses money and revise its entries, exits, sizing, and regime/noise handling using chronological, arm-attributed evidence. Keep the paper book adequately funded for the experi…
work: CURRENT STATE 2026-09-01T15:25Z, replacing a note that read as resolved. THE EVIDENCE PROGRAMME THIS ITEM DEPENDS ON IS FAILING ON TWO FRONTS. Epoch thesis-action-v2-same-row-20260815 is VOID-IMMINENT: 423 development clusters collected but the leading proposal-eligible cell holds 1 of 20, because the protocol admits o…
latest substantive update: WORK ACTIVE_IMPLEMENTATION: CURRENT STATE 2026-09-01T15:25Z, replacing a note that read as resolved. THE EVIDENCE PROGRAMME THIS ITEM DEPENDS ON IS FAILING ON TWO FRONTS. Epoch thesis-action-v2-same-row-20260815 is VOID-IMMINENT: 423 development clusters collected but the leading proposal-eligible cell holds 1 of 20, b…
P1 CURRENT OWNER: PLATFORM OI-172 · 7d old
EVIDENCE-FIRST RUNTIME EVOLUTION: establish a reproducible clean deployment identity, one validation-first command/service capability manifest, independent offsite restore proof, canonical agent-control parity, Node/Linux/Windows compatibility evidence, and a …
work: refreshed 2026-08-31: operator ruled four buttons 2026-08-30 19:30 MST - G4 shaping caps adopted, drawdown bar wide at 5000bps, spend asks priced, G3 import delegated
latest substantive update: REVIEW: G1 universe-replay audit cut fix committed on branch review at 108acccd (one instant cut for all ledgers, at-cut terminals reported never refused, 6/6 tests, live dry run REPLAY_MATCHED at 05:47Z where the old cut refused at 05:40Z; the audit CLI is a pinned manifest module so its row and the core constants are…
P1 CURRENT OWNER: PLATFORM OI-178 · 3d old
`pump-creates.log.jsonl` crossed Node's string ceiling and crashed `holderTrajectoryCli` 12 times, but its writer now rotates at 150 MB, archives before rotation, and has fired in production; its three known whole-file consumers stream and the 90% warn tier no…
work: NEXT: lane-executable residuals remain after the rotation fix - unread archives and the thirteen unaudited artifacts are ours to work, not blocked
latest substantive update: WORK ACTIVE_IMPLEMENTATION: NEXT: lane-executable residuals remain after the rotation fix - unread archives and the thirteen unaudited artifacts are ours to work, not blocked
P1 CURRENT OWNER: PLATFORM OI-180 · 3d old
MONITORING THAT CANNOT SEE WHAT IT CERTIFIES (R3, R8, R9, R13, R14, R28, R34): exit-invariant watch rests on a false liveness premise (only the runner emits an unconditional counters line; established's sits inside a branch and its mtime heartbeat is fed by th…
work: NEXT FALSIFICATION: spawn the real CLI post-fix with exact STOPPED markers, no receipts and no cursor and require exit 2; then prove durable cursor persistence across two invocations; then repair the causal-cap control, which a whole-file-read-then-slice passed while reading 16416 bytes
latest substantive update: WORK ACTIVE_IMPLEMENTATION: NEXT FALSIFICATION: spawn the real CLI post-fix with exact STOPPED markers, no receipts and no cursor and require exit 2; then prove durable cursor persistence across two invocations; then repair the causal-cap control, which a whole-file-read-then-slice passed while reading 16416 bytes
P1 CURRENT OWNER: REVIEW OI-184 · 3d old
CLI smoke already spawns 11 money-path-shaped CLIs, including all three paper engines, for up to 20 seconds inside a temporary-cwd sandbox; the exact-production interrupted-pass/restart-idempotency harness is still unbuilt, excluded daemons and `.cmd` wrappers…
work: REVIEW 2026-09-02T19:00Z: fuller ask designed on the board. Under the lock the engine appends the book row, then the sidecar event, then the evidence journal and positions.set; three kill cells (before the book append; between book and sidecar; after the sidecar) on the real engine through the wrapper's shape, trigger …
latest substantive update: WORK ACTIVE_IMPLEMENTATION: REVIEW 2026-09-02T19:00Z: fuller ask designed on the board. Under the lock the engine appends the book row, then the sidecar event, then the evidence journal and positions.set; three kill cells (before the book append; between book and sidecar; after the sidecar) on the real engine through t…
P1 CURRENT OWNER: REVIEW OI-198 · 47h 2m old
GATE-RECHECK RELIABILITY: fib-feature times out on 210 of its 270 executions (78%) and prebond-shape on 210 of 383 (55%), each failure freezing that gate for 6h, so accruing-gate evidence is mostly stale
work: REVIEW 2026-09-02T09:08Z CAUSE MEASURED: the fib-feature gate's working set reaches 3.2 GB (it materializes the 1.53 GB curve-outcomes tape as full rows for three fields); joining all 257 attempts since 08-19 to runner-death-capture freemem samples: 217 kills at median 1.9 GB free (83% under 4 GB), 40 oks at median 4.4…
latest substantive update: NIGHTLY RED again 7:42 AM at 0cfc6f5f on the same sampler control. Both reds (7:00, 7:42) sit inside windows where PLATFORM's own test suites held free memory at 1.6 to 2.4 GB; green at 5:53 on a quieter host. No sampler change until it fails on a quiet host; PLATFORM stops daytime nightly re-arms.
P2 CURRENT OWNER: PLATFORM OI-204 · 42h 16m old
Both freshness guards are wired after the watchdog mutex and their receipts are age-checked, but the scheduled context uses labelled noncanonical Node 24 because canonical Node 22 is absent from its visible AppData\Local namespace; `alertDigest` still does not…
work: POST-9732701c5 CORRECTION TO THE IMMEDIATELY PRIOR CODEX UPDATE: PLATFORM landed a consumer-side three-hour receipt-age gate and stamped both guard receipts with nodeVersion/nodeExec while that update was being filed. Therefore the earlier statement that alertDigest had no age gate is now stale. Canonical Node-22 recei…
latest substantive update: TWO FACTUAL CORRECTIONS to the bounded-exposure note. The conclusion is unchanged; two of the statements supporting it were wrong. 1. "data/sol-thesis-peak-state.jsonl ABSENT" is FALSE. The file exists and is the largest artifact on this desk: data/receipts/thesis-action/v2-same-row-20260815/sol-thesis-peak-state.json…
P3 CURRENT OWNER: PLATFORM OI-189 · 2d old
COMMIT PERIMETER CAN HARD-BLOCK BOTH LANES, AND THE ONLY ESCAPE DROPS IT ENTIRELY (Review Two N2, N3, N4, N30, N31). All shipped 2026-08-30. N2: plainLanguageGate scans the whole staged diff with NO pathspec, and escalation.ts:1727,1847 writes two banned words…
work: commit-perimeter work continuing; stale openItemsCli path in the refusal text repaired 5fbae512 and the scan widened 68c137e6
latest substantive update: WORK ACTIVE_IMPLEMENTATION: commit-perimeter work continuing; stale openItemsCli path in the refusal text repaired 5fbae512 and the scan widened 68c137e6

RESEARCH 12

collecting evidence or narrowing the diagnosis
P1 CURRENT OWNER: REVIEW OI-182 · 3d old
Mark preparation is no longer independently replayed by four producers: `ba04d7ff3` centralizes preparation and append with controlled costs of unchanged `4S`, mark advance `4S+2delta`, replacement `3S0+4S1`, and financial advance `3S0+delta+4S1`; final prefix…
work: REVIEW 2026-09-02T09:30Z MEASURED: the ba04d7ff seam is active (both engines restarted after it). Runner since 02:55Z restart: 276 LOCK SUMMARY windows, 0 refusals (the established sleeve is idle, zero lock lines since its restart), hold max per window p50 996 ms, p90 7.3 s, max 48.5 s, 197 hold warnings. Driving the e…
latest substantive update: REVIEW 2026-09-02T15:38Z CORRECTION: PLATFORM's confound holds. The first suite-free hour (15Z, 28 windows) reads hold max p50 136 ms / p90 1.6 s / max 7.8 s, back at the floor's first-hour level; the 13Z-14Z worsening (p50 1.5 s, max 10-13 s) was PLATFORM's seven clean-worktree suite runs evicting the ledger's pages, …
P1 CURRENT OWNER: PLATFORM OI-219 · 6h 17m old
solTapeV2's pump leg (data/sol-tape.pump.jsonl, the birth tape) has written nothing since 2026-09-02T04:18Z (9:18 PM MST 09-01): totals pump=1306 frozen across 2,294 of the daemon's own silence reconnects while the pumpswap leg wrote 1.5M rows; the watchdog's …
work: Next: recycle solTapeV2 (standing capture authority, pid receipts) as the first test; if pump rows do not resume within 10 min of a fresh subscription, the WS logs filter or the upstream pump program is the cause and it goes to code. Then: the producer-silence alert needs a consumer (a bounded auto-recycle or a board E…
latest substantive update: SUPERVISOR CLOSED 10:26 AM (a6e77511): run-sol-tape2.cmd is in watchdog.ps1's reboot-survival list with a data/SOL-TAPE2-STOPPED marker gate; pinned by tests/watchdogSupervisesTapeDaemon; active at the next tick. Still open: the birth-floor decision (operator), a consumer for the producer-silence alert (operator's choi…
P2 CURRENT OWNER: PLATFORM OI-171 · 8d old
COINGECKO ANALYST CREDIT SPRINT, ENDED: the paid-host routing sprint and its temporary override are over as of the September 1 replenishment boundary; what remains is evaluating adoption and outcomes from the forward route/credit/outcome receipts the sprint ca…
work: EXPIRED OPERATOR-ONLY OVERRIDE CLEARED, 2026-09-01T07:02Z. Jeremy's overnight instruction to keep the lanes working and get the gates green supplied the operator authority the prior blocker required. `.env` now contains no `COINGECKO_HARD_STOP_FRAC` line, so the code's unchanged 0.85 default is authoritative again. Can…
latest substantive update: 2026-09-02T05:25Z PLATFORM RECEIPT: ruling applied. .env GT_PAID_BULK_UNTIL 2026-09-01 -> 2026-10-01T00:00:00.000Z; showdown recycled (wrapper 90044, node 90928, 10:01 PM MST; a one-minute duplicate from the watchdog relaunch gap was retired). Meter 05:12Z: bulk->paid, 52 paid ok, 0 429, 0 free-by-policy, credits 794/5…
P2 CURRENT OWNER: PLATFORM OI-185 · 3d old
The 2.4 GB `held-price-path.jsonl` and companion held-mark tape are frozen at 2026-08-13; the three behaviorally checked consumers expose freshness or UNKNOWN rather than false current marks, but the peak-reachability instrument is dark for newer experiments, …
work: NEXT: only 3 of 37 consumers checked; 34 remain plus the successor-wiring versus retire decision
latest substantive update: PRIORITY P1 -> P2: Only 3 of 37 consumers have been measured and the remaining slice is bounded research/retirement routing. Preserve the stale-reader risk, but rank it below measured gate timeouts and direct book/causal identity defects. [checked-one-file-is-not-checked]
P2 CURRENT OWNER: PLATFORM OI-217 · 12h 4m old
A runner-v1 position (pool 2Lcp1M9s, mint 7VwjRc8i) was held 61 minutes with no mark row at all, no refusal line in the engine log, and then exited on a hard stop at 1:20 AM MST with -0.049 SOL realized: the exit price came from somewhere other than a booked m…
work: NEXT: read the runner engine's mark-request selection (which positions the safety pass asks to mark and why one per pass); establish what price the hard stop used; determine whether the engine's open-position view and the ledger projection disagree by one. Consequence not yet measured: how many exits since 08-14 fired …
latest substantive update: FIX LANDED 08acbdec (3:35 AM 09-02): lastMarkTs written only by booked marks; selectMarkCandidates keys on it in all three sleeves. NOT ACTIVE until the runner and established sleeves restart (morning item 4, with OI-196).
P2 CURRENT OWNER: PLATFORM OI-220 · 4h 9m old
solShadowPriceMarkCli ran 52 five-minute cycles (4:53 to 9:12 AM 09-02) with status NO-RPC and wrote no shadow mark: it had no dotenv load of its own and inherited one through its encodeBase58 import from pumpCreatesCli, which the OI-197 landing (de521f7c, 4:5…
work: Fix: import dotenv/config at the top of solShadowPriceMarkCli (in the tree now); sweep every src/cli/*Cli.ts for the same shape and pin it with a test; the daemon picks the fix up on its own next loop, no recycle.
latest substantive update: REVIEW 2026-09-02T16:52Z: the class is live on the money path, not one refactor away. Import-graph census (480 CLIs, de521f7c^ vs master): 80 CLIs reached a dotenv load only through pumpSwapPoolMint > pumpCreatesCli and reach none now; seven read secrets by name, the three sleeve engines among them; none imports paperR…
P3 CURRENT OWNER: PLATFORM OI-173 · 3d old
CG WebSocket G3 (OnchainOHLCV) push-marks trial: replace polled ohlcv for held pools with a streamed subscription at 0.1 credit/response, gated on a measured response rate. ohlcv is the largest paid line item on the CG key (showdown alone: 45,041 paid requests…
work: reliability alarm withdrawn - it was the probe, not CoinGecko; the billing-versus-reliability hose decision is still open
latest substantive update: WORK ACTIVE_RESEARCH: reliability alarm withdrawn - it was the probe, not CoinGecko; the billing-versus-reliability hose decision is still open
P3 CURRENT OWNER: PLATFORM OI-174 · 3d old
OI-148 second-tier CG candidates have no adoption owner: megafilter (diamond #2, the band screener with honeypot/sell-tax/FDV filters matching the 100k-2M charter), coins/top_gainers_losers as a regime feature, coins/list/new graduation watch, and a coins/sola…
work: EXPIRED OPERATOR-ONLY OVERRIDE CLEARED, 2026-09-01T07:02Z. Under Jeremy's overnight operator instruction, the time-boxed `.env` `COINGECKO_HARD_STOP_FRAC=0.98` line was removed after its September 1 boundary. The unchanged code default is 0.85; canonical Node v22.23.2 expiry controls pass 2/2. Fresh process-table attri…
latest substantive update: WORK ACTIVE_RESEARCH: EXPIRED OPERATOR-ONLY OVERRIDE CLEARED, 2026-09-01T07:02Z. Under Jeremy's overnight operator instruction, the time-boxed `.env` `COINGECKO_HARD_STOP_FRAC=0.98` line was removed after its September 1 boundary. The unchanged code default is 0.85; canonical Node v22.23.2 expiry controls pass 2/2. Fre…
P3 CURRENT OWNER: REVIEW OI-190 · 2d old
PLATFORM'S OWN NEW INSTRUMENTS SHIPPED CARRYING THE DEFECT CLASS THEY POLICE (Review Two N11, N24, N28, N32, N35, N38, N39). N24: hookLivenessCheck.mjs:127 hardcodes <root>/.git/hooks - the identical resolution bug its sibling test had been repaired the SAME D…
work: N38 partition withdrawn as exact after CODEX rejected my decomposition; remaining points still under audit
latest substantive update: REVIEW 2026-09-02T16:15Z: the sidecar-rule instance is fixed on review@316b743e (conflict classes on what the id does not bind, a sixth control that can fail, coverage counter); live PASS 255/72/0 with six controls; mutation matrix over the real 72 pairs flips every pair per perturbed field. Landing asked. Filed instan…
P3 CURRENT OWNER: PLATFORM OI-191 · 2d old
THE COINGECKO RUNWAY ALARM CANNOT FIRE, AND FIVE NEW CAPTURES IGNORE THE BUDGET (Review Two N1 critical, N17, N18, N19, N21, N22). N1: escalation.ts:1736 raises the alarm only when runway < daysLeftInWindow, and runway divides remaining credits by a rate ancho…
work: source completion recorded; the status-page observation was point-in-time only and retry/billing semantics remain open
latest substantive update: WORK ACTIVE_RESEARCH: source completion recorded; the status-page observation was point-in-time only and retry/billing semantics remain open
P3 CURRENT OWNER: PLATFORM OI-194 · 2d old
A DEAD 17-DAY-OLD STREAM IS REFERENCED BY 65 FILES, THREE OF THEM ON THE WATCHDOG CADENCE. data/candidate-trajectory.log.jsonl last accepted a write 2026-08-13T21:08:34Z and is a frozen 574MB file. The exposure is curveRunoutCli, dashboardCli and showdownCli; …
work: NEXT: the filing lane explicitly refused to close on a partial fold; unexamined consumers and the reframed scope remain
latest substantive update: WORK ACTIVE_RESEARCH: NEXT: the filing lane explicitly refused to close on a partial fold; unexamined consumers and the reframed scope remain
P3 CURRENT OWNER: PLATFORM OI-201 · 44h 14m old
Later-gate readiness evidence is neither absent nor safely auto-mintable: most rows carry imported evidence and many legitimately REFUSE; the open defect is expiry, re-import, and accountable authorship, while EXTERNAL_EVIDENCE and RUNTIME_DEPLOYED rows must r…
work: NEXT: the 27 later-gate checks are enumerated and their 18 refusals split 9 stale / 9 current; the open question is which of the 9 stale verdicts change on a re-import, which requires re-running those audits rather than re-stamping them
latest substantive update: 24 VERSUS 25 RECONCILED, and both counts are right about different questions. CODEX posted that the imported-evidence partition is 25 of 27, not the 24 I recorded. Folding the newest receipt for both predicates separately: later-gate rows 27 state !== UNKNOWN …

VALIDATION 3

reviewed work awaiting its named receipt
P1 CURRENT OWNER: REVIEW OI-183 · 3d old
BUY entries still model against observed rather than decision-time reserves, but reserve age is now recorded and buys more than 30 minutes stale refuse; a one-day post-recycle sample measured p50 6.06 seconds and `-0.0550 SOL` (`-3.757%` of 1.463 SOL) across 1…
work: REVIEW 2026-09-02T18:36Z: mechanism answered. Decomposition of decisionAt minus reserveObservedAt over 77 runner entries joined to the sidecar: observed-to-available 0 on all; the median 6.3 s (floor 5.35 s) sits inside the engine pass: 4.3 s from availability to the mint-authority read (the loop sleeps 5 s, drains the…
latest substantive update: REVIEW 2026-09-02T20:15Z: fix 1 committed on review@f58ed4ec on the clean base 89cdc713 (gate PASS, no override): the context carries the vault read's clock, source and provenance; older rows validate; no fill changes; latencyMs becomes the true age. Landing asked; activation at the sleeve restart. Fix 2 (peek under th…
P1 CURRENT OWNER: REVIEW OI-196 · 2d old
LOCK RELEASE CAN PERMANENTLY WEDGE A LIVE ENGINE: both shared-ledger and with-lock release guards disable retry before unlink succeeds; a Windows EBUSY leaves a same-live-PID residue peers cannot reclaim.
work: REVIEW 2026-09-02T08:50Z: fix on branch review at 3cce5862 (src/data/arm1SharedLedgerLock.ts, scripts/with-lock.mjs, tests): release retries a delete denial over 25/50/100/200/400 ms, never throws, records and names a persistent denial; a process removes its own post-release residue at its next acquisition (nested hold…
latest substantive update: NIGHTLY RED 3:46 AM 09-02 at e37e948e: sharedLedgerLockRace one-holder case, 2/3 late racers receipted inside 15 s; passes alone (19/19, 24 s); free memory 6.2-9.9 GB during the run; second red of this test in the log.
P3 CURRENT OWNER: REVIEW OI-197 · 2d old
Seven base58 encoders diverge: three throw SyntaxError on an empty buffer where four return empty string, and fixed-offset Buffer.slice clamps to empty on a short account
work: REVIEW 2026-09-02T11:15Z: steps 1-3 on review@4f3d8fad: src/core/base58.ts is the one body, seven encoder bodies removed (pumpCreatesCli and poolPriceAdjudicatorCli re-export it under their old names), nine importers repointed to the leaf, vectors pinned (empty, leading zeros, all-zero key, PumpSwap id, random round tr…
latest substantive update: WORK REVIEW_COMPLETE: REVIEW 2026-09-02T11:15Z: steps 1-3 on review@4f3d8fad: src/core/base58.ts is the one body, seven encoder bodies removed (pumpCreatesCli and poolPriceAdjudicatorCli re-export it under their old names), nine importers repointed to the leaf, vectors pinned (empty, leading zeros, all-zero key, PumpSw…

BLOCKED 8

a named dependency prevents the next action
P1 CURRENT OWNER: PLATFORM OI-177 · 3d old
Local hook enforcement is no longer dead: the installed pre-commit hook is BINDING, helper-shaped line-wise reads are detected for added staged lines, and watchdog liveness exposes absent/drifted hooks and stale/failing CI; at last measurement scheduled CI sti…
work: BLOCKER: the enumerated gate-0 residuals need operator dirty-tree cleanup and a canonical-skills direction ruling; neither is a lane action
latest substantive update: PRIORITY P0 -> P1: The original P0 premise that the guard binds nowhere was refuted: the installed pre-commit hook matches the tracked guard. The residual dirty-tree and canonical-skills authority dependencies remain real, but they no longer outrank the desk's direct book/causal-integrity P0. [self-report-is-not-proof]
P1 CURRENT OWNER: REVIEW OI-181 · 3d old
OI-181 WATCHDOG MUTEX STILL SPANS 35 DIRECT POTENTIALLY BLOCKING CALLS; TIMEOUT SAFETY REMAINS UNKNOWN. Commit ba04d7ff3 adds a fail-closed controller-v2 STARTED/VERIFIED authority, but the running watchdog has not loaded it and no natural deployed receipt exi…
work: BLOCKER: source/test repair is pushed at ba04d7ff3 (18/18 focused, 120/120 architecture, both typechecks), but the running watchdog has not loaded it. Controlled activation, natural controller-v2 STARTED/VERIFIED evidence, and the retained 35-child deployment boundary require separate operator-authorized runtime work; …
latest substantive update: NIGHTLY RED 3:46 AM 09-02 at e37e948e: watchdogCiObserverDeadline blocked-gh case, controller did not close in 43 s (ran 53 s); passes alone (18/18, 60 s); free memory 6.2-9.9 GB during the run; second red of this test in the log.
P1 CURRENT OWNER: PLATFORM OI-195 · 2d old
OI-195 INTRA-FUNCTION REVIEW CLASSIFIES 36/36 DECLARED ROWS AS RACE-CAPABLE; THE REAL LEGACY BOARDPOST PATH IS RED AND THE CRASH-RELEASED TARGET IS GREEN, WHILE CALLER REACHABILITY, EXTERNAL SERIALIZATION, RUNTIME FIRING, DAMAGE, FAMILY MIGRATION, AND ACTIVATI…
work: Source target, discriminator, family census, and the b23c0e650 immutable rebind are committed and independently clean-verified at 502771715. Remaining activation requires an operator-authorized fenced disable/drain/residue-inspection/generation-flip/restart window plus natural real-entrypoint acceptance. No recycle or …
latest substantive update: OWNER ABSORBED CODEX -> PLATFORM by scripts/ownerAbsorptionCheck.mjs --absorb. CODEX is not in the active lane set (PLATFORM) Work does not fall off a dark lane; the active set holds it until the original owner returns and reclaims it.
P2 CURRENT OWNER: PLATFORM OI-205 · 41h 56m old
Two lanes on ONE working tree interfere in both directions: a lane's mid-edit state fails the other lane's whole-project commit gate, and a pathspec commit swallows the other lane's dirty hunks in any shared file
work: Operator dependency, not executable queue work: choose per-lane worktree isolation or a staged-snapshot gate repair, then supply an acceptance control proving staged-bad/worktree-good and omitted untracked tests cannot pass. Until that choice, shared-tree interference cannot be safely implemented away by either lane. […
latest substantive update: THE ORPHAN MECHANISM IS NOW MEASURED, not inferred, and the obvious candidate is REFUTED. I said the next step was new instrumentation to record who holds the lock at orphan time. It was not needed twice over: the existing log answered the attribution question, and two direct experiments answered the mechanism questio…
P2 CURRENT OWNER: PLATFORM OI-206 · 39h 13m old
25 of 36 readiness evidence rows are past their own expiresAt and 10 of those are STILL MARKED VERIFIED (worst 58.2h overdue); accountable lanes PLATFORM 18 / CODEX 7, and the count has been growing - 15/36 with 4 VERIFIED when this was filed
work: Operator/external and clean-HEAD dependencies make the current next step non-executable: 13 of 25 expired rows require external attestation, 2 require deployed-runtime evidence, and the 10 nominally recollectable committed-HEAD rows still require a clean fixed snapshot. Preserve REFUSE/UNKNOWN rather than treating this…
latest substantive update: 2026-09-01T23:05Z: the daily readiness receipt task (scripts/run-readiness-receipt-daily.cmd) pinned the Node 22 runtime at a path its scheduled context could not see (MSIX virtualization, see OI-213), and its isolated worktree lives in the same invisible root. Seven consecutive non-SUFFICIENT attempts and the 25/36 ex…
P3 CURRENT OWNER: PLATFORM OI-186 · 3d old
The legacy real-money leaf entry point was deleted at `79d878738`; R17/R18 remain only in the stopped legacy paper runner—ratchet pending state is erased on HOLD and the two agnostic knobs are present but nonbinding—with fix-or-delete and historical impact unr…
work: BLOCKER: the remaining R17/R18 placebo-knob choice is operator authority, not a lane call; the deleted entry point closed only R19 reachability
latest substantive update: PRIORITY P1 -> P3: The reachable real-money entry point was deleted; this paper-only desk still has a residual future-config/placebo-knob policy question, but it is operator-blocked and not a current money-path exposure. [scope-is-the-instructions-words]
P3 CURRENT OWNER: REVIEW OI-202 · 43h 25m old
214 mark rows (lower bound at 2026-09-01T01:08:15Z) were written to the money-path ledger stamped rpcVaultMarkLineageStatus VERIFIED_V5 under the reverted equal-dual rule; production CONTAINED at 01:08:34Z by recycling the stale-code runner
work: BLOCKER: containment and the consumer census are DONE (runner recycled 01:08:34Z, one gating fold identified and never-run); the only remaining action is whether 214 append-only money-path rows are corrected, annotated or left as history, which is operator authority and not a lane call
latest substantive update: REVIEW 2026-09-02T12:50Z: whole-ledger fold bounds the laundered set exactly: 214 VERIFIED_V5 runner marks in the equal-dual window (00:41:57Z to 01:07:48Z 09-01, 10 entries, 214 receipts), none after containment until OI-188 restored legitimate marks at 09-02T02h. The vault-mark receipts carry no rule marker (same sev…
P3 CURRENT OWNER: PLATFORM OI-203 · 43h 8m old
Append-only headline corrections are implemented and centrally resolved by the CLI, dashboard source, and evidence gateway; runtime activation remains unaccepted until a fresh served render proves the active dashboard worker resolves a newly corrected face.
work: BLOCKER: source and fresh-CLI behaviour are verified (3bded1896 contract, 9ad89762d wiring + 8 spawned-CLI tests, applied to OI-194 and OI-202), so QUEUED was measured-false. What remains is activation: the served dashboard worker predates the source change and still renders the old headline, and tsx does not hot-reloa…
latest substantive update: CODEX DOCUMENTATION-GATE REPAIR, commit 80de61e9: PLATFORM commit 531e0985c added the shared src/core/openItemHeadline.ts resolution boundary and rewired the CLI, dashboard, and evidence gateway, but named it in none of the durable architecture documents. TARGET_PROCESS_MAPS now records immutable what plus append-only …

QUEUED 8

not active now; waiting for its next work slice
P1 CURRENT OWNER: PLATFORM OI-170 · 10d old
PumpSwap has a verified partition cutover, but OI-170's causal-boundary manifest remains review-only: module-level reachability yields two false indirect writer routes, one real live Runner writer route, and one mutable replay-reader blocker.
work: 2026-09-01T23:45Z: operator chose in chat (~4:33 PM MST, button 'Snapshot into git') to commit CODEX's unfinished third repair whole; landed as a75c9a86 (10 files, gate green) on top of 7a6c1ed4 (the test slice). HEAD is one coherent state again. Next for PLATFORM: read the third repair from a75c9a86 against CODEX's la…
latest substantive update: REVIEW: the nine Gate 0 fast-suite reds in the OI-170 files (7 hermeticity sites, 2 vacuity sites) are green in the clean worktree at review@7165b9b5, tests only; awaiting PLATFORM landing and the leg's re-collection. Slice 2 landed on master as fa24a09b (byte-identical to 316aa7be).
P1 CURRENT OWNER: PLATFORM OI-192 · 2d old
SHARED-TREE DURABILITY AND HYGIENE GAPS EXPOSED BY TWO LANES AT SPEED (Review Two N25, N26, N29, N33, N34). N26: data/board/MARKETDATA.events.jsonl is the only lane board tape NOT in git - not ignored, simply never added - so a shipping lane's entire board his…
work: no work recorded and no dated update; shared-tree durability gaps are scoped but untouched
latest substantive update: PRIORITY P3 -> P1: Repeated shared-index absorption plus the fresh surface audit confirm the durability/hygiene objective remains untouched while its failure class is actively recurring. Promote from P3 so MARKETDATA tape durability, untracked-hook visibility, exact shared-tree hygiene, CRLF policy, and containment cov…
P1 CURRENT OWNER: PLATFORM OI-210 · 31h 41m old
THE NIGHTLY CLEAN-WORKTREE GATE OMITS THE TWO TSC ARMS, so the one instrument that judges committed code cannot see a typecheck break: HEAD could not compile its own test suite for ~26h while every commit-gate run reported green off each lane's dirty tree - th…
work: DEMONSTRATED, not theorised, 2026-09-01 ~05:45 MST. CODEX ran the full suite in an isolated CLEAN WORKTREE for 5c07f5dc and it failed on something neither lane's gate has ever reported: HEAD:src/data/arm1Ledger.ts:611 function appendMarkLedgerRows(...) <- no `export` HEAD:tests/arm1PreLockMintObser…
latest substantive update: CORRECTION TO MY OWN CORRECTION: I named the wrong driver. The conclusion is unchanged and better sourced. I wrote in the previous note that Task Scheduler drives the nightly gate via "coldread-gate-recheck, next run 2026-09-01 06:31 MST". That is false. I matched a task by name without reading its action. coldread…
P2 CURRENT OWNER: PLATFORM OI-193 · 2d old
NOTHING VERIFIES THAT A COMMIT CLAIMING CLOSED ACTUALLY CLOSED ANYTHING (Review Two N27, N40). N40: openItemCitationCheck verifies only that the cited id EXISTS; no check requires a commit claiming CLOSED to stage the corresponding registry transition. This is…
work: no work recorded and no dated update; close-verification gap is scoped but untouched
latest substantive update: CORRECTION: the 2026-09-01T08:33Z claim that OI-193 was the only substantively untouched queued objective was false. OI-192 also remains scoped-but-untouched on each named repair surface. The honest untouched-objective count is two: OI-192 and OI-193. OI-184 and OI-187 mention a specific harness/ratchet not started, bu…
P2 CURRENT OWNER: PLATFORM OI-215 · 14h 38m old
SOL held marks are single-witness by construction and no decision recorded it: the two-source price-integrity quarantine belonged to the USD paper book and closed with its 08-10 archive; the 08-13 rpc-vault move left SOL held marks with one witness. Operator d…
work: Waits on the operator's morning choice between (a) and (b), presented as a button with the price of each; PLATFORM builds (a) on approval. Evidence is OBSERVED: REVIEW's OI-175 trace (launcher exits on the USD archive marker, last held-mark row 2026-08-14T04:08Z, verify class zero since 08-27); the consequence of a sin…
latest substantive update: WORK QUEUED: Waits on the operator's morning choice between (a) and (b), presented as a button with the price of each; PLATFORM builds (a) on approval. Evidence is OBSERVED: REVIEW's OI-175 trace (launcher exits on the USD archive marker, last held-mark row 2026-08-14T04:08Z, verify class zero since 08-27); the consequ…
P3 CURRENT OWNER: PLATFORM OI-187 · 3d old
HYGIENE, SPRAWL AND GUARD-TIGHTENING BATCH (R21, R22, R23, R24, R25, R26, R27, R29, R30, R31, R32, R33, R35): paid-host boundary is convention-only with no containment test; Triton partition cutover wedged since Aug 24 with writers routed back to the 827MB mon…
work: NEXT: the guard-tightening ratchet is PROPOSED, NOT BUILT, and its own update says why it was not started
latest substantive update: WORK QUEUED: NEXT: the guard-tightening ratchet is PROPOSED, NOT BUILT, and its own update says why it was not started
P3 CURRENT OWNER: REVIEW OI-214 · 16h 27m old
OI-170 causal reader-census verdict has no reader and no index-resolution guard: a pinned-module edit flips it to MANIFEST_MISMATCH silently
work: NEXT (slice 3 of the OI-170 finish, REVIEW): (1) a scheduled reader of the census verdict: the nightly clean-worktree gate (or the watchdog tick) runs `solCausalBoundaryManifestCli --census-perimeter` in the readiness-clean checkout, records status + manifestIssues count + scannedFiles in its receipt, and pages on any …
latest substantive update: REVIEW slice 3b committed at review@56cfd4a2 (on master 54201616): scripts/manifestPinCheck.mjs as the tenth pre-commit gate; a staged edit to a pinned module refuses unless the re-pinned row, the staged manifest and the core's two digest constants are in the same commit; 5/5 scratch-repo tests plus a size-guarded real…
P3 CURRENT OWNER: PLATFORM OI-221 · 54m old
A stale git temporary index .git/next-index-70008.lock (279 KB, DIRC v2 image from 2026-08-17 16:13 local, the temp index git writes during stash and split) sits in the main repository with no holder. REVIEW classified it removable; PLATFORM's delete was decli…
work: Next action is the operator's hand: delete C:\Users\Jeremy\Cwod\bankr-robinhood-agent\.git\next-index-70008.lock when no git process is running (16 days old). REVIEW: the next-index-<pid>.lock shape is a candidate class for gitLockCensus.
latest substantive update: WORK QUEUED: Next action is the operator's hand: delete C:\Users\Jeremy\Cwod\bankr-robinhood-agent\.git\next-index-70008.lock when no git process is running (16 days old). REVIEW: the next-index-<pid>.lock shape is a candidate class for gitLockCensus.

UNCLASSIFIED 0

work state missing or unrecognized; not evidence that work never started
none

READY TO CLOSE 0

never inferred from prose
The canonical open-item writer has no READY_TO_CLOSE state. This dashboard deliberately shows zero rather than guessing from a review note; close only through the writer once the stated acceptance evidence exists.

RECENTLY CLOSED 5

archive tail, not deleted work
CLOSED OI-208 · 2h 20m ago
Resolved in two steps on branch review, both landed by PLATFORM. Step one (e0a35830, 0efd3f37): the board/git lock-order inversion reproduced with the real applyRotation and the real with-lock on temp roots, both wait edges witnessed at one instant, bounded in…
CLOSED OI-216 · 3h 9m ago
Resolved by the untouched-holdout audit instrument (untouchedHoldoutAuditCli, landed 1e1ff91d-era on master, --candidate at bb22cf3b): binds the selection receipt's generation and digest, reproduces the protocol's train/holdout partition from the receipt's own…
CLOSED OI-218 · 4h 41m ago
Both rows have committed producers with --write --candidate: quantified-completeness (d9e19031, evidence row 31 REFUSE on two feeds, honest) and decision-quote-versus-executable-fill (8d509791, row 33 VERIFIED, PASS on 255 fills; sidecar rule brought to the wr…
CLOSED OI-179 · 11h 21m ago
REVIEW 2026-09-02T09:10Z CLOSED with PLATFORM's no-objection (DM 08:59Z). Receipts: whole-ledger fold of 131,466 lines / 11,060 booking rows on the book-wide bankroll and cumulative-realized chains, 0 breaks, so the steal race never produced a fork; the serial…
CLOSED OI-175 · 14h 46m ago
CLOSED as answered by REVIEW 2026-09-02T05:40Z on PLATFORM's ACK (DM 05:38Z). The three questions the item asked: (1) accounting: the verify class (CoinGecko cross-check leg of held-position marks) ran 679 to 2,589 attempts a day through 08-09, fell to 449 on …

Evidence & control detail

blindspot, confidence, and policy-selection context
BLINDSPOT → OI 1 actionable · 2 linked OI · 62 collecting evidence · 32 benign · 32 legacy awaiting current classification
Validated findings enter this OI ledger immediately. Underweight findings retain a named evidence gap; benign findings do not inflate the backlog.
SOL CONFIDENCE CLAIMS v2 18d ago - Phase 2 confidence-weighted-fresh-sol-paper - claims/promotion/rollback only
PAPER AUTHORIZATION Phase 2 - sol-phase-2-exploratory-paper-authorization-v1 - 3 active sleeve(s), exact 1.0x exploratory paper
sol-migration-entry-v2 - AUTHORIZED 1.0x
sol-established-dip-buy-v2 - AUTHORIZED 1.0x
sol-runner-entry-v2 - AUTHORIZED 1.0x
TARGET ONLY EXCLUDED sol-birth-entry-v2 - sol-dip-buy-v2 - sol-runner-continuation-entry-v3 - sol-runner-recycle-entry-v3
ACTIVE RUNTIME sol-migration-entry-v2
none / UNKNOWN promotion NO; rollback NO; never a paper-size control (strategy-specific observation count unknown; strategy-specific coverage unknown; no immutable heldout/forward validation receipt)
ACTIVE RUNTIME sol-established-dip-buy-v2
none / UNKNOWN promotion NO; rollback NO; never a paper-size control (strategy-specific observation count unknown; strategy-specific coverage unknown; no immutable heldout/forward validation receipt)
ACTIVE RUNTIME sol-runner-entry-v2
none / UNKNOWN promotion NO; rollback NO; never a paper-size control (strategy-specific observation count unknown; strategy-specific coverage unknown; no immutable heldout/forward validation receipt)
POLICY SELECTION: REFUSE 49m ago · 2743 paired trips · 31 candidates · 100% receipt coverage · 100% exact causal authorization
baseline trail-25/30; receipt evidence since 2026-08-26 15:00:00Z; holdout receipts 0%; holdout causal authorization 0%; holdout stress 0%; no challenger clears every promotion gate. This governs promotion/rollback claims only; it does not alter Phase-2 exploratory paper sizing.
Close with --close <id> --note "<how>" --lesson "<what changes>". Closed items move to data/closed-items-archive.jsonl, not deleted.

Window
this book (live) all history
showing 2026-08-09 08:51Z · 1:51 AM → now · 11082 rows (11082 executed · 0 shadow)
5 · ARCHIVE & RAW

Books & raw system facts
THE BOOK (SOL, three sleeves): -240.09% · -240.0941 SOL realized on 100 SOL · 135292 rows

Archived USD book

FROZEN 2026-08-10
The USD paper book was archived by operator ruling on 2026-08-10 (paperRunner stopped). The five boards this card replaces — 1 Trade learning, 2 Off-book learning, 3 Theories, 4 Applied changes, and the USD-era Instruments daemons — spent ~19% of this page on a book that will not advance. Nothing was deleted: the evidence stays on disk and in git history.
data/trade-autopsy.jsonl · data/paper-trades.log.jsonl · data/active-experiments.json · data/experiment-queue.json · data/retired-experiments.jsonl · data/autopilot-actions.log.jsonl · data/promoted-knobs.json · data/paper-summary.json · data/held-stream-marks.json · data/missed-opportunity-ledger.jsonl · data/entry-fillability-shadow.log.jsonl · data/curve-graduation-summary.json
Live-book equivalents: grading — Oracle & exit diagnosis · producer liveness — Learning loop pulse · experiments — Experiment scoreboard. (Curve graduation dynamics was retired separately: 11.7d-stale manual artifact with no runner and no machine consumer.)

Trade history
1 historical unmatched exit(s) remain quarantined (OI-128, already corrected) and are excluded from this table. They do not hide the rest of the book.
newest 100 of 11082 rows shown — 10982 older omitted so the page stays loadable · show newest 600 · 11082 executed (5545 buys · 5537 sells) + 0 shadow (ERROR / RISK_BLOCKED) in window · 11082 rows in file · tap a row for detail
timetradesizePnLPnL %
09-02
20:24Z · 1:24 PM		 Hero SOL Ck377NeTAQ1ayhBtNjm3umscNkpekBtcJHjPJrtNpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
19:45Z · 12:45 PM		 Hero SOL Ck377NeTAQ1ayhBtNjm3umscNkpekBtcJHjPJrtNpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · trail		 0.10 SOL ~-0.00076 SOL -0.8%
no detail fields on this row
09-02
19:42Z · 12:42 PM		 Hero SOL Ck377NeTAQ1ayhBtNjm3umscNkpekBtcJHjPJrtNpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
19:06Z · 12:06 PM		 8Bit SOL 4XbRwc96Bd4QbMMepT27Naz95GFUCoXd7KoUfgy9pump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.10 SOL ~-0.0038 SOL -3.8%
no detail fields on this row
09-02
18:45Z · 11:45 AM		 DEG SOL 7HJCxCMTnit9rLS21e9SMdisJaaZ9A5P3N1Lq99roeTz
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · trail		 0.10 SOL -0.01 SOL -11.7%
no detail fields on this row
09-02
17:23Z · 10:23 AM		 UNDEADS SOL EguENXBejC9EvHqqfKZ1koVnUNbbYM1EcLsPcJtxpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
17:16Z · 10:16 AM		 $LEMSICK SOL 2YAY8PucaY9zVFvyUArjMB12HxtRRrrTtzLRuwpLpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
17:12Z · 10:12 AM		 DEG SOL 7HJCxCMTnit9rLS21e9SMdisJaaZ9A5P3N1Lq99roeTz
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
16:43Z · 9:43 AM		 UNDEADS SOL EguENXBejC9EvHqqfKZ1koVnUNbbYM1EcLsPcJtxpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · trail		 0.10 SOL -0.04 SOL -43.8%
no detail fields on this row
09-02
16:40Z · 9:40 AM		 UNDEADS SOL EguENXBejC9EvHqqfKZ1koVnUNbbYM1EcLsPcJtxpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
16:35Z · 9:35 AM		 OHNO SOL 8vhDy8y77v8z6NRoWfkQF1e4QEcgwMd4zZmGeWqhpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.09 SOL ~-0.0034 SOL -4.0%
no detail fields on this row
09-02
16:33Z · 9:33 AM		 DEG SOL 7HJCxCMTnit9rLS21e9SMdisJaaZ9A5P3N1Lq99roeTz
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.10 SOL -0.05 SOL -55.5%
no detail fields on this row
09-02
16:31Z · 9:31 AM		 DEG SOL 7HJCxCMTnit9rLS21e9SMdisJaaZ9A5P3N1Lq99roeTz
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
16:27Z · 9:27 AM		 NORMIE SOL 2mAiYQgxpBn8ztj36gm7sT94SqUCfUa85MaaeYD9eRHW
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.10 SOL -0.02 SOL -18.4%
no detail fields on this row
09-02
16:16Z · 9:16 AM		 Bros SOL 75etQwhSzicjbDNK2V7LEEEdj422S9HMwTraWR4Dpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.06 SOL — —
no detail fields on this row
09-02
15:37Z · 8:37 AM		 FAMI SOL 92brNaWExrd7FWX4F6ij64Zcc5VathbNJU53hXd2in5u
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
14:54Z · 7:54 AM		 pippo SOL Gffw364rz4r93aYum3BHynoi5iw1gsq2m4P2Py6gpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.10 SOL -0.05 SOL -51.5%
no detail fields on this row
09-02
13:52Z · 6:52 AM		 GAJN SOL 5BDz2cdep7FjXYhk473qDV2MCBw4dF7jVBg3Mcrnpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.10 SOL -0.01 SOL -8.8%
no detail fields on this row
09-02
13:29Z · 6:29 AM		 Gronk SOL 2TesPAqafiad1bKac6AxWSEXD4JnUP6AzapBz3L3pump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.08 SOL ~-0.0040 SOL -5.1%
no detail fields on this row
09-02
13:24Z · 6:24 AM		 pippo SOL FDaGGXRv8qxuZoDFnkcGCkRMexXWKyrErYtB2b8vaiqB
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
13:13Z · 6:13 AM		 pippo SOL Gffw364rz4r93aYum3BHynoi5iw1gsq2m4P2Py6gpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
11:23Z · 4:23 AM		 ☉ SOL GyaBdGkrktJdJJcqoXi7mVC37d8eECu5SetPz8TSpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.10 SOL -0.05 SOL -51.3%
no detail fields on this row
09-02
10:47Z · 3:47 AM		 Biu SOL 4rxNSU5BynixWPZvN3Nprvm4fdomjRSQynFb5XvQpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.10 SOL -0.05 SOL -51.1%
no detail fields on this row
09-02
10:10Z · 3:10 AM		 ☉ SOL GyaBdGkrktJdJJcqoXi7mVC37d8eECu5SetPz8TSpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
10:08Z · 3:08 AM		 Biu SOL 4rxNSU5BynixWPZvN3Nprvm4fdomjRSQynFb5XvQpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
10:03Z · 3:03 AM		 JIM SOL KUiToAcZ981sNmJ85Y9Ah7oiLsNo7p24nEoWbvnpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.10 SOL -0.01 SOL -12.3%
no detail fields on this row
09-02
09:13Z · 2:13 AM		 Booby SOL GDTj9R5XoBSVEUQzJXL8R8aZoojo3gPJxkHfeZqzpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
09:07Z · 2:07 AM		 Ducky SOL CyJdFagKW4xPMy6EPKRNy2GcJkebpBKJfu3YCyubpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.10 SOL -0.01 SOL -7.4%
no detail fields on this row
09-02
08:20Z · 1:20 AM		 Bros SOL 7VwjRc8it6s3J19NFvc31To7NZvfJjiHvAM2z9KTpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.10 SOL -0.05 SOL -50.3%
no detail fields on this row
09-02
07:21Z · 12:21 AM		 MOTION SOL Biq6zwK4sRB9DtXajmXfo5a5X7B7FWvDN2h38M8mpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.10 SOL ~-0.0040 SOL -4.0%
no detail fields on this row
09-02
07:17Z · 12:17 AM		 Bros SOL 7VwjRc8it6s3J19NFvc31To7NZvfJjiHvAM2z9KTpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
07:06Z · 12:06 AM		 8Bit SOL 4XbRwc96Bd4QbMMepT27Naz95GFUCoXd7KoUfgy9pump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
06:55Z · 11:55 PM		 Hashcash SOL AL63Th2HZAjHzU5hvRiqsDumfYp3CEnZUQaFSDnVotQn
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · trail		 0.10 SOL +0.06 SOL +57.6%
no detail fields on this row
09-02
06:51Z · 11:51 PM		 Pill SOL 4232dGVrUST2LsZVJweoGmSmeqBtQdRe25aEUm3jb43E
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · trail		 0.10 SOL +0.01 SOL +10.9%
no detail fields on this row
09-02
06:42Z · 11:42 PM		 Hashcash SOL AL63Th2HZAjHzU5hvRiqsDumfYp3CEnZUQaFSDnVotQn
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
06:29Z · 11:29 PM		 Uptember SOL U29rFVTgvGpYuKRe2YYSkjzNkz9J1D5CCeUfVGopump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.10 SOL +~0.00071 SOL +0.7%
no detail fields on this row
09-02
06:04Z · 11:04 PM		 Hashcash SOL AL63Th2HZAjHzU5hvRiqsDumfYp3CEnZUQaFSDnVotQn
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.10 SOL -0.05 SOL -50.3%
no detail fields on this row
09-02
05:59Z · 10:59 PM		 Hashcash SOL AL63Th2HZAjHzU5hvRiqsDumfYp3CEnZUQaFSDnVotQn
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
05:59Z · 10:59 PM		 alon SOL 72NtVkCkBC2mKCT6paAdo9g4KptUrRH8aRU3xzBVQHDW
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.10 SOL -0.05 SOL -50.0%
no detail fields on this row
09-02
04:35Z · 9:35 PM		 OHNO SOL 8vhDy8y77v8z6NRoWfkQF1e4QEcgwMd4zZmGeWqhpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.09 SOL — —
no detail fields on this row
09-02
04:33Z · 9:33 PM		 CLAYNO SOL GZtHTMrxs3BKExgRQEriRcSX4NZNemBzkTF4vw8b6EYc
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.08 SOL ~-0.0017 SOL -2.1%
no detail fields on this row
09-02
04:27Z · 9:27 PM		 NORMIE SOL 2mAiYQgxpBn8ztj36gm7sT94SqUCfUa85MaaeYD9eRHW
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
04:27Z · 9:27 PM		 Pill SOL 4232dGVrUST2LsZVJweoGmSmeqBtQdRe25aEUm3jb43E
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
02:59Z · 7:59 PM		 Pumpcat SOL 8y16oHsekNSNM38A6ZkN9yTJGQyFyye71P8y1KHopump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.10 SOL -0.05 SOL -50.7%
no detail fields on this row
09-02
02:49Z · 7:49 PM		 Pumpcat SOL 8y16oHsekNSNM38A6ZkN9yTJGQyFyye71P8y1KHopump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
02:28Z · 7:28 PM		 alon SOL 72NtVkCkBC2mKCT6paAdo9g4KptUrRH8aRU3xzBVQHDW
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
02:23Z · 7:23 PM		 PIG SOL BwpwLTytiF6zshcuxb7yrfEY2BCiRaAcLaxuDcccBqTm
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · trail		 0.10 SOL +0.01 SOL +11.8%
no detail fields on this row
09-02
01:51Z · 6:51 PM		 GAJN SOL 5BDz2cdep7FjXYhk473qDV2MCBw4dF7jVBg3Mcrnpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
01:48Z · 6:48 PM		 DoodleCat SOL 9Yo4GnDHVfZVNhni7VspHroTsfVfrXfEqR3nkQjvh3xT
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · trail		 0.10 SOL -0.01 SOL -5.4%
no detail fields on this row
09-02
01:47Z · 6:47 PM		 DoodleCat SOL 9Yo4GnDHVfZVNhni7VspHroTsfVfrXfEqR3nkQjvh3xT
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
01:38Z · 6:38 PM		 PIG SOL BwpwLTytiF6zshcuxb7yrfEY2BCiRaAcLaxuDcccBqTm
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
01:29Z · 6:29 PM		 Gronk SOL 2TesPAqafiad1bKac6AxWSEXD4JnUP6AzapBz3L3pump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.08 SOL — —
no detail fields on this row
09-02
01:28Z · 6:28 PM		 MENCY SOL ARFoxh9owBGns33xh9sFddH9V3zzsiXiRj3wuUCkpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.06 SOL ~-0.0033 SOL -5.2%
no detail fields on this row
09-02
01:02Z · 6:02 PM		 084BEAR SOL 4eXGQJMBZp656x9eqerEc1hepSoe27zvuXBtNuqjpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.10 SOL ~-0.0013 SOL -1.3%
no detail fields on this row
09-02
00:50Z · 5:50 PM		 PIG SOL BwpwLTytiF6zshcuxb7yrfEY2BCiRaAcLaxuDcccBqTm
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.10 SOL -0.05 SOL -50.5%
no detail fields on this row
09-02
00:33Z · 5:33 PM		 PIG SOL BwpwLTytiF6zshcuxb7yrfEY2BCiRaAcLaxuDcccBqTm
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-02
00:33Z · 5:33 PM		 JELITICS SOL 4rRDyAMWWnegXxCHD1qLipEfHVxktr8HeEta9Y2epump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.10 SOL -0.01 SOL -6.5%
no detail fields on this row
09-01
23:57Z · 4:57 PM		 FishPro SOL 7A7PAR65fZ9BCCRk8E2k5AthbFPJjqsaEDw7yH5apump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.11 SOL -0.03 SOL -26.0%
no detail fields on this row
09-01
23:51Z · 4:51 PM		 PIG SOL BwpwLTytiF6zshcuxb7yrfEY2BCiRaAcLaxuDcccBqTm
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · trail		 0.10 SOL +0.01 SOL +12.2%
no detail fields on this row
09-01
23:29Z · 4:29 PM		 PIG SOL BwpwLTytiF6zshcuxb7yrfEY2BCiRaAcLaxuDcccBqTm
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-01
23:19Z · 4:19 PM		 Rusty SOL PrqsV8SUtYXnZfzqS7jcVG1ExaUNJULuS7fN4XSpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.11 SOL -0.01 SOL -10.2%
no detail fields on this row
09-01
22:03Z · 3:03 PM		 JIM SOL KUiToAcZ981sNmJ85Y9Ah7oiLsNo7p24nEoWbvnpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-01
21:49Z · 2:49 PM		 PPPP SOL GNB9MA4a3VmnJPNSEwU8awFyqpYDnY9A257axnCpug87
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.11 SOL -0.03 SOL -25.4%
no detail fields on this row
09-01
21:07Z · 2:07 PM		 Ducky SOL CyJdFagKW4xPMy6EPKRNy2GcJkebpBKJfu3YCyubpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-01
21:00Z · 2:00 PM		 Little SOL FyM68CzuG7cSJ9Ao2JAZt1uQSVrj9SRyrnVAVnJ7pump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.10 SOL -0.05 SOL -53.7%
no detail fields on this row
09-01
20:05Z · 1:05 PM		 Little SOL FyM68CzuG7cSJ9Ao2JAZt1uQSVrj9SRyrnVAVnJ7pump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-01
19:56Z · 12:56 PM		 DICKBUTT SOL CgypxZcmmZMVQMERA6D59hkh8maPEnj31pYRKM9cpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · trail		 0.10 SOL +0.03 SOL +31.7%
no detail fields on this row
09-01
19:21Z · 12:21 PM		 MOTION SOL Biq6zwK4sRB9DtXajmXfo5a5X7B7FWvDN2h38M8mpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-01
19:21Z · 12:21 PM		 DICKBUTT SOL CgypxZcmmZMVQMERA6D59hkh8maPEnj31pYRKM9cpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-01
19:03Z · 12:03 PM		 Pumpatar SOL EcdsHpecmRLq3tFaTSCfkn4WojdeMskNa695QW81g7Y4
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.10 SOL -0.06 SOL -58.9%
no detail fields on this row
09-01
19:00Z · 12:00 PM		 1% SOL EGTa8RQ5rnK9sSd1xkVJPA4wAQgY5jjhtxCwZYrDpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.06 SOL -0.02 SOL -34.0%
no detail fields on this row
09-01
18:29Z · 11:29 AM		 Uptember SOL U29rFVTgvGpYuKRe2YYSkjzNkz9J1D5CCeUfVGopump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-01
18:26Z · 11:26 AM		 ADLANA SOL 4p2WVwtMBY2poBPCUiDf7wPGWPLjG393gdG9iTugpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · trail		 0.10 SOL -0.01 SOL -14.2%
no detail fields on this row
09-01
17:57Z · 10:57 AM		 Pumpatar SOL EcdsHpecmRLq3tFaTSCfkn4WojdeMskNa695QW81g7Y4
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-01
17:40Z · 10:40 AM		 GROK SOL t8Nzr5jGUzQjHm8GxX4F2AmiB4CX38moRm8UrFkpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.11 SOL -0.02 SOL -20.4%
no detail fields on this row
09-01
16:32Z · 9:32 AM		 CLAYNO SOL GZtHTMrxs3BKExgRQEriRcSX4NZNemBzkTF4vw8b6EYc
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.08 SOL — —
no detail fields on this row
09-01
16:28Z · 9:28 AM		 nekomaru SOL 6JhHZqZYJ7w4G86CQxHZy1ajQcDoNxNhsPrCfHLwpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.09 SOL ~-0.0024 SOL -2.7%
no detail fields on this row
09-01
14:24Z · 7:24 AM		 ADLANA SOL 4p2WVwtMBY2poBPCUiDf7wPGWPLjG393gdG9iTugpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-01
14:23Z · 7:23 AM		 CHUCHO SOL Daj8qEnEb9L6v1wmrCPJGunJAsQJuXctFsEh9TsApump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.11 SOL -0.06 SOL -54.1%
no detail fields on this row
09-01
13:28Z · 6:28 AM		 MENCY SOL ARFoxh9owBGns33xh9sFddH9V3zzsiXiRj3wuUCkpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.06 SOL — —
no detail fields on this row
09-01
13:24Z · 6:24 AM		 NFA SOL 5ffneJV96Q3X9jeuJTPajPSZxCWVFdssWrVTTy11pump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.10 SOL -0.05 SOL -51.3%
no detail fields on this row
09-01
13:02Z · 6:02 AM		 084BEAR SOL 4eXGQJMBZp656x9eqerEc1hepSoe27zvuXBtNuqjpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-01
12:50Z · 5:50 AM		 NFA SOL 5ffneJV96Q3X9jeuJTPajPSZxCWVFdssWrVTTy11pump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-01
12:32Z · 5:32 AM		 JELITICS SOL 4rRDyAMWWnegXxCHD1qLipEfHVxktr8HeEta9Y2epump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-01
12:13Z · 5:13 AM		 084BEAR SOL 4eXGQJMBZp656x9eqerEc1hepSoe27zvuXBtNuqjpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.10 SOL -0.06 SOL -56.9%
no detail fields on this row
09-01
12:13Z · 5:13 AM		 084BEAR SOL 4eXGQJMBZp656x9eqerEc1hepSoe27zvuXBtNuqjpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.10 SOL — —
no detail fields on this row
09-01
11:57Z · 4:57 AM		 FishPro SOL 7A7PAR65fZ9BCCRk8E2k5AthbFPJjqsaEDw7yH5apump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.11 SOL — —
no detail fields on this row
09-01
11:25Z · 4:25 AM		 CHUCHO SOL Daj8qEnEb9L6v1wmrCPJGunJAsQJuXctFsEh9TsApump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.11 SOL — —
no detail fields on this row
09-01
11:19Z · 4:19 AM		 Rusty SOL PrqsV8SUtYXnZfzqS7jcVG1ExaUNJULuS7fN4XSpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.11 SOL — —
no detail fields on this row
09-01
10:53Z · 3:53 AM		 orangie SOL eoAHM5xZeT1VwivG1CKusHJeuPHzpyUKRa8PywCpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.11 SOL ~-0.0019 SOL -1.8%
no detail fields on this row
09-01
10:44Z · 3:44 AM		 Rusty SOL PrqsV8SUtYXnZfzqS7jcVG1ExaUNJULuS7fN4XSpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · trail		 0.11 SOL +0.02 SOL +20.9%
no detail fields on this row
09-01
10:30Z · 3:30 AM		 LUCKLAY SOL 7rWzV3T1Sge8xEQW45aYCeoJVS8aUABiyMThkafBqqwG
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.11 SOL -0.01 SOL -5.5%
no detail fields on this row
09-01
09:49Z · 2:49 AM		 PPPP SOL GNB9MA4a3VmnJPNSEwU8awFyqpYDnY9A257axnCpug87
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.11 SOL — —
no detail fields on this row
09-01
09:45Z · 2:45 AM		 084BEAR SOL 4eXGQJMBZp656x9eqerEc1hepSoe27zvuXBtNuqjpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · hard-stop		 0.11 SOL -0.06 SOL -52.2%
no detail fields on this row
09-01
09:34Z · 2:34 AM		 084BEAR SOL 4eXGQJMBZp656x9eqerEc1hepSoe27zvuXBtNuqjpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.11 SOL — —
no detail fields on this row
09-01
09:02Z · 2:02 AM		 084BEAR SOL 4eXGQJMBZp656x9eqerEc1hepSoe27zvuXBtNuqjpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · trail		 0.10 SOL ~-0.0041 SOL -4.0%
no detail fields on this row
09-01
08:56Z · 1:56 AM		 JAILSTOOL SOL 2xMkwGXieXBXwDo7EBJwzmhqn4vRPCwifVqf2i13pump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.06 SOL ~-0.00062 SOL -1.1%
no detail fields on this row
09-01
08:53Z · 1:53 AM		 Rusty SOL PrqsV8SUtYXnZfzqS7jcVG1ExaUNJULuS7fN4XSpump
BUY arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner		 0.11 SOL — —
no detail fields on this row
09-01
08:51Z · 1:51 AM		 STOOLPRESIDEN SOL HsPFjCz9uNk9JXY9ao6GZ42PkNDf3KsKcxp26jgDpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.11 SOL -0.02 SOL -16.9%
no detail fields on this row
09-01
08:34Z · 1:34 AM		 PRIME SOL CWfkh3BddLbSoVt2adcvewofiyYevcfXrCiU1iGcpump
SELL arm1 · Runner sleeve · sol-runner-entry-v2 · raw arm=runner-v1 · raw entryRule=runner · time-stop		 0.06 SOL ~-0.00066 SOL -1.1%
no detail fields on this row
generated 2026-09-02 20:26Z · 1:26 PM · data snapshot 0s old · built in 1998ms
page age 0s · auto-refresh 90s (in place — the page is not reloaded)
clocks: UTC · MST (UTC-7) · naive filter times read as UTC
COLDREAD v2 · LIVE SOL BOOK · THREE SLEEVES