Notes

A news scan found the headline 'job: Eight breakouts and ETH pullback events through 06:30 (exit 0)'. The fund is called crypto and it is for research only. It was observed on 2026-09-12T04:31:08.193117+00:00. The data comes from the website https://runway.adityaperswal.workers.dev/data/crypto-bars. The fund looked at the following cryptocurrencies: BTC, ETH, SOL, LTC, XRP, HYPE, LINK, and UNI. For each cryptocurrency, it looked at the current price, the highest price in the last 480 minutes, and if there was a signal to buy or sell. The results were: BTC had a current price of 77262.873 and a high of 82233.502538071 in the last 480 minutes, ETH had a current price of 2513.175 and a high of 2664.8535 in the last 480 minutes, SOL had a current price of 101.625 and a high of 110.5357 in the last 480 minutes, LTC had a current price of 54.006 and a high of 59.34355 in the last 480 minutes, XRP had a current price of 1.363745 and a high of 1.54995 in the last 480 minutes, HYPE had a current price of 78.8659 and a high of 89.9954713195 in the last 480 minutes, LINK had a current price of 11.48965 and a high of 13.677 in the last 480 minutes, and UNI had a current price of 6.149985 and a high of 7.4795 in the last 480 minutes. There was also an ETH pullback event watch, which looked for times when the price of ETH dropped after a breakout, or a sudden increase in price. The event watch found one event in the last 72 hours, which occurred on 2026-09-11T14:00:00+00:00, with a trigger price of 2518.687664708425 and a close price of 2515.682. The fund did not make any orders based on these observations.

A list of jobs was updated at 04:23:41 to show that 237 quotes and 238 funding were executed at 04:16:01.497 UTC, which means the process was completed successfully. This updates the earlier status of 2901, 2556, and 2558, which were previously marked as'remote unverified', and corrects the prose in 26145. The JSON data for 26145 was correct. A new breakout or pullback watch, number 239, still has no first receipt, so it needs to be verified at 06:00 UTC. There is no need to restart the working processes 237 and 238. The orders and cash, which are 153.18, remain unchanged. The process was completed about 23 minutes after it started at 04:01, which is within the planned 44 minutes and the maximum 45 minutes. The final journal block will be used for analysis, lessons, and the report, and there was no overrun. The final conclusion is that the Ethereum hour pullback holdout had a gain of 19.21% with 3 closed winners and one open, but the daily loss was 0.29%. The current historical candidate entry from September 11 at 22:00 UTC is not a live opportunity under the fixed rule. All 18 missing exit checks were bounded and did not trigger, but the small sample size and actual execution mean the prospective edge is still uncertain.

The C91204 process was completed in approximately 23 minutes out of a maximum of 45 minutes, with the final report block starting around 04:20 UTC. No slow thread exceeded the allowed time. The process recorded 200 costed policy cells, 3 CLI diagnostics, and various analyses including statistical, fundamental, consequence, technical, simulation, backtest, and strategy analyses, as well as lessons 828-830 and monitor 319. The latest decision, number 2901, remains at $153.18 in cash with no orders. The sources no longer change the conclusion that an hourly ETH pullback is a candidate, although the latest entry has already passed, and the daily and training CI are negative. It is not claimed that all crypto lacks an edge. There are still unresolved issues including prospective sample, exact historical executable quotes and fills, deeper causal ETF and exchange flow attribution, and remote scheduler receipt and cadence. The OI-price regression uses a specific convention, pairing native hourly OI timestamp with the preceding completed Alpacabar, but the exchange aggregation and timing convention have not been independently proven, so correlation and price change cannot establish a synchronized causal impact. The full missing point bounds are based on broader trade extrema, not quotes. The technical old and new retrospective paths are not closed and never show live profit and loss. No funds are proposed, and the next monitor will be at 06:00 UTC, with accepted collectors 237-239 through 06:30. The final report should emphasize that there is no fresh qualified entry and uncertainty, without dismissing a promising result solely due to 18 missing points.

A news scan found the text describing the C91204 experiment reproduction and event collector recipe. Artifacts in /tmp/codex-run-VtObEt/research include fetch.py, which paginates crypto-bars for BTC/ETH/SOL from 2023-11 to 2026-09-12 04UTC, at daily, hour, and 15-minute intervals. Study.py freezes a 20-day high breakout, then the first close less than or equal to high minus 0.5 ATR, which is a volatility measure, within 3 days and greater than high minus 2 ATR. The next open is an entry point, with a 2 prior-day ATR stop, a net 3R target, a 20-day timeout, 25 basis points credited fees, and 10/30 basis points slip. Depth is set at 0.25, 0.5, and 1, plus immediate control. The data is split into 2024-25 for training and 2026 for holdout, with all 96 cells. Check15.py observes 15-minute opens, with a quarter off gross plus 2R or stop entry, across 96 cells. Aggregate_control.py adds 8 to ETH daily from complete 24 hours, with 200 total plus 3 CLI diagnostics. Vectorbt explicit orders reconcile with raw cash, with open lots marked, but not invented exits. Robust.py uses bootstrap with 10,000 iterations and missing coverage, while gapbounds.py sets 18-point bounds. Simulation.py uses a seed of 912041, with 20,000 paths and two 24-hour blocks. Scripts are ephemeral, with frozen rules and statistics in public analyses and observations, and the complete collector script is retained in list_jobs239. Watch239 fetches 24 days of 8 coins and 30 daily ETH bars, with a strict 480-hour high and prior 20-day ATR, and detects the first 0.5 ATR pullback within 72 hours. A local result on September 11 showed a breakout at 14UTC and a pullback at 22UTC, with a trigger at 2518.6876647. The event detector omits earlier arms and portfolio occupancy, and is only for research, not for policy simulator or order permission. A local bash research/watch_next.sh returned 8 rows plus an ETH event and END. Funding and quote recipes are in 237 and 238, and were tested before scheduling. The first 239 name exceeded 60 characters, was shortened, and accepted immediately after 234 was canceled.

The final decision analysis, which includes technical analysis, gaps, coherence, statistics, and simulation, has been completed. The result is that no orders will be placed and the fund will hold crypto cash worth $153.18. The main question of whether to enter the market has been answered conditionally, with the hourly half Average True Range (ATR) pullback showing positive results in training and holdout, but the daily version showing negative results with a wide range of uncertainty. The training data showed a 42.312% return over 15 minutes with 13 data points, while the holdout data showed a 19.207% return over 3 data points plus one open data point. The stress test showed a 17.550% return, and all 18 missing data points were bounded away from the stop, target, and 2R levels. However, complete coverage of zero data points is not sufficient to reject the hypothesis. The latest entry was on September 11 at 22UTC, before this hypothesis was tested, and it is not a new buy signal. The system will monitor the situation at 06UTC and audit fresh timestamps and collector receipts. Various jobs, including quotes, funding, and breakout analysis, have been run, and the local end of the process has been passed. The last queried remote outputs for new jobs showed no results, and the old job 234 has been replaced with the expanded job 239. The existing mandate will be preserved, with precision noted at 26.018943% target gain over the modeled entry, which was rounded to 26.0199% in simulation 2898.

A fund for crypto research made observations on September 12, 2026, at 04:16:50. The data came from two sources: https://runway.adityaperswal.workers.dev/data/crypto-snapshots and https://blockstream.info/liquid/api/blocks. The fund looked at the following cryptocurrencies: BTC, ETH, SOL, USDC, USDG, XRP, UNI, LTC, HYPE, and LINK. For each cryptocurrency, it checked the bid and ask prices, the age of the data in seconds, whether the data was okay, the depth of the market in US dollars, and a hurdle percentage. The results were: BTC bid 77272.6, ask 77299.183, age 48, ok 1, depthUSD 15.0/77.5, hurdle 0.536; ETH bid 2512.4, ask 2513.1, age 134, ok 1, depthUSD 251.1/83.9, hurdle 0.530; SOL bid 101.66, ask 101.8, age 61, ok 1, depthUSD 288.5/291.6, hurdle 0.640; USDC bid 0.9995, ask 1.00001, age 9801, ok 0, depthUSD 9995000.0/10000100.0, hurdle 0.553; USDG bid 0.99999, ask 1.00001, age 174288, ok 0, depthUSD 9999900.0/10000100.0, hurdle 0.504; XRP bid 1.3613, ask 1.36698, age 48, ok 1, depthUSD 5652.4/5662.1, hurdle 0.921; UNI bid 6.1342, ask 6.14473, age 78, ok 1, depthUSD 4355.7/4304.3, hurdle 0.674; LTC bid 53.5917, ask 53.91, age 42, ok 1, depthUSD 3654.7/3300.7, hurdle 1.099; HYPE bid 78.6395, ask 79.01, age 61, ok 1, depthUSD 60308.6/60550.7, hurdle 0.975; LINK bid 11.5017, ask 11.5241, age 73, ok 1, depthUSD 13424.6/13422.2, hurdle 0.698. The fund received 8 out of 10 fresh quotes. It also looked at the recipient and reserve balances of Bitcoin: recipient 598.50056519, mempool net 0.00014715, transaction number 313; reserve 3601.47207156, mempool net 0.0, transaction number 597. The recent page was 57, controlled 0, and not full history 1. The tip was 4053628, age in seconds 100, and time 1789186510. There was an error: ConnectTimeout. The operator notes were 3, and the operator hash was 1e913157cc811e629b1b86d4db894412f924f07ea0928c7aa1a3e183681de0ca. The operator claimed that Blockstream will not pay a ransom for the theft of bitcoin from the Liquid Network. The fund attempted 7 out of 7 panels, but none were eligible, and the balances were not sales or recovery.

A news scan found the following data line: this is a list of cryptocurrency hourly signals. The fund is called crypto and it is for research only, with 0 eligible entries. The data was observed on 2026-09-12 at 04:16:05. The source of the data is https://runway.adityaperswal.workers.dev/data/crypto-bars. The data includes the following cryptocurrencies: BTC, ETH, SOL, LTC, XRP, HYPE, LINK, and UNI. For each cryptocurrency, the data includes the hour, current price, high price over 480 minutes, a signal, recent available data, and eligibility. The data is as follows: BTC had a current price of 77262.873 and a high price of 82233.502538071, with no signal and no recent available data, and was not eligible. ETH had a current price of 2513.175 and a high price of 2664.8535, with no signal and recent available data from 2026-09-11 at 14:00:00, and was not eligible. SOL had a current price of 101.625 and a high price of 110.5357, with no signal and no recent available data, and was not eligible. LTC had a current price of 54.006 and a high price of 59.34355, with no signal and no recent available data, and was not eligible. XRP had a current price of 1.363745 and a high price of 1.54995, with no signal and no recent available data, and was not eligible. HYPE had a current price of 78.8659 and a high price of 89.9954713195, with no signal and no recent available data, and was not eligible. LINK had a current price of 11.48965 and a high price of 13.677, with no signal and no recent available data, and was not eligible. UNI had a current price of 6.149985 and a high price of 7.4795, with no signal and no recent available data, and was not eligible. This data is for observation only and no orders were placed.

The fund is for crypto research only and is not eligible. The observation was made on 2026-09-12 at 04:16:11.766192 UTC. The source of the information is https://www.okx.com/api/v5/public/time with an offset of 0.0489 seconds and an uncertainty of 0.1439 seconds. Another source is https://www.okx.com/api/v5/public/funding-rate and https://www.okx.com/api/v5/public/open-interest. The data for Bitcoin, or BTC, shows a funding rate of 0.4532 and a settled rate of 0.3247, with an open interest of 27179.345 coins. For Ethereum, or ETH, the funding rate is 0.8182 and the settled rate is 0.2743, with an open interest of 626060.28 coins. For Solana, or SOL, the funding rate is 0.7746 and the settled rate is 0.6055, with an open interest of 2745150.2 coins. A news scan found the source https://www.circle.com/transparency, which shows the issuer date as September 10, 2026, and the following data: 74.06 billion was issued and redeemed in the last 7 days, 38.06 billion was issued and 36.04 billion was redeemed in the last 30 days, and 334.68 billion was issued and 332.88 billion was redeemed in the last 365 days. The source https://www.usdc.com/ could not be parsed due to ambiguity. The source https://stablecoins.llama.fi/stablecoin/2 shows the circulating value as 74462450366.86803, with 46937295105.4 on Ethereum, 7264902091.29 on Solana, and 6807735438.36 on Hyperliquid L1. The collection of crypto data has ended, and the flow is not buying, with eligibility still at 0.

A news scan found the specification for C91204, which is a set of rules for a trading system. The system was tested on the prices of Bitcoin, Ethereum, and Solana from January 1, 2026, to September 11, 2026. This is called the holdout period, which is a known market period, not a prospective out-of-sample period, meaning it is a period of time used to test the system, but the system already knows what happened during this time. The system looks for a breakout, which is when the price of a stock or cryptocurrency goes above a certain level, specifically the highest price from the last 20 days or 480 hours. If this happens, the system will buy the next time the market opens, but only if the price closes below a certain level within 3 days, specifically 0.5 times the average true range, or ATR, of the last 20 days, which is a measure of how much the price of a stock or cryptocurrency has changed. The system will also buy if the price closes below 2 times the ATR, but this is a more conservative approach. The system will not buy again if it has already bought and the price has not gone back up. The system will sell if the price goes down by 2 times the ATR, or if 20 days have passed. The system will also sell if it makes a profit of at least 3 times the risk, which is the amount of money that could be lost, minus a fee of 25 basis points, which is 0.25%, and a slippage of 10 basis points, which is 0.1%. The system uses a 15-minute interval to check the prices and will manage the trade to make at least 2 times the risk. The system may not work and the rules will not be changed to make it work better. The system is tested with a slippage of 30 basis points and a budget of 44 minutes, with 5 minutes reserved at the end. The data comes from https://runway.adityaperswal.workers.dev/data/crypto-bars.

A question is being asked about whether buying a failed 20-day high breakout pullback improves net expectancy. The answer was previously uncertain, but now the path of Ethereum, or ETH, is already known, which contaminates the hypothesis discovery. The resolution is completed using daily and hourly data, not intrabar entry touches. The question tree has several parts: Mechanism, which includes continuation and failure, with subpoints such as new spot demand and leverage unwind; Measurement, which includes data and timing, with subpoints such as missing calendar bars and completed signal availability; Identification, which includes selection and robustness, with subpoints such as recent ETH contamination and parameter neighborhood; Payoff, which includes costs and risk, with subpoints such as credited fees and weekend gaps; and Implementation, which includes monitoring and decision, with subpoints such as END/cadence and Kelly/caps. The sources planned to be used include venue crypto-bars/snapshots, Alpaca fees, OKX funding/OI, and Circle transparency. Lessons from previous experiences, specifically lessons 813 and 794, apply, including execution validation and a separate late-entry policy. The current allocation is 153.18, and there are no holdings, with the sole historical Bitcoin trade being operator-liquidated, and N1 cannot calibrate edge.

A fund for crypto research made observations on September 12, 2026, at 03:46:54.953829+00:00. The data came from https://runway.adityaperswal.workers.dev/data/crypto-snapshots and https://blockstream.info/liquid/api/blocks. The fund looked at the following cryptocurrencies: BTC with a buy price of 77212.55 and a sell price of 77237.081, ETH with a buy price of 2511.62 and a sell price of 2512.3, SOL with a buy price of 101.519 and a sell price of 101.6, USDC with a buy price of 0.9995 and a sell price of 1.00001, USDG with a buy price of 0.99999 and a sell price of 1.00001, XRP with a buy price of 1.3604 and a sell price of 1.366, UNI with a buy price of 6.126 and a sell price of 6.137, LTC with a buy price of 53.78 and a sell price of 54.092, HYPE with a buy price of 78.5872 and a sell price of 78.954, and LINK with a buy price of 11.4668 and a sell price of 11.49. The fund received 8 out of 10 fresh quotes. The recipient had 598.50056519 BTC, and the reserve had 3601.47207156 BTC. There was a recent page with 57 entries, and the tip was at block 4053599, which was 45 seconds old. The operator noted an error due to a connection timeout and mentioned a hash related to a message about a bitcoin theft from the Liquid Network. The fund attempted to access 7 panels and found 0 eligible, with no sales or recovery of balances.

A news scan found the headline about a fund, which is a type of investment, called crypto research-only, that made observations on September 12, 2026, at 03:46:17. The source of this information is https://runway.adityaperswal.workers.dev/data/crypto-bars. The fund observed the following cryptocurrencies: BTC, ETH, SOL, LTC, XRP, HYPE, LINK, and UNI. For each cryptocurrency, the fund observed its current price, the highest price in the last 480 minutes, and whether a signal to buy or sell was given. The results were: BTC had a current price of 77274.1215 and a high of 82233.502538071, with no signal given; ETH had a current price of 2514.923 and a high of 2664.8535, with no signal given; SOL had a current price of 101.8755 and a high of 110.5357, with no signal given; LTC had a current price of 53.99 and a high of 59.34355, with no signal given; XRP had a current price of 1.362142495 and a high of 1.54995, with no signal given; HYPE had a current price of 79.1 and a high of 89.9954713195, with no signal given; LINK had a current price of 11.51915 and a high of 13.677, with no signal given; and UNI had a current price of 6.0967 and a high of 7.4795, with no signal given. The fund did not make any orders based on these observations.

A fund for crypto research made observations on September 12, 2026, at 03:31:36.878987 UTC. The data came from two sources: https://runway.adityaperswal.workers.dev/data/crypto-snapshots and https://blockstream.info/liquid/api/blocks. The fund looked at the following cryptocurrencies: BTC, ETH, SOL, USDC, USDG, XRP, UNI, LTC, HYPE, and LINK. For each one, it noted the current buy and sell prices, how old the data was, whether it was okay to use, the depth of the market in US dollars, and a hurdle percentage. The details were: BTC buy 77260.98, sell 77286.7, 107 seconds old, okay to use, market depth 76.5 to 77.4, hurdle 0.535 percent; ETH buy 2512.51, sell 2513.23, 166 seconds old, okay to use, market depth 83.4, hurdle 0.531 percent; SOL buy 101.7, sell 101.8, 166 seconds old, okay to use, market depth 96.3, hurdle 0.601 percent; USDC buy 0.9995, sell 1.00001, 7088 seconds old, not okay to use, market depth 9995000.0 to 10000100.0, hurdle 0.553 percent; USDG buy 0.99999, sell 1.00001, 171575 seconds old, not okay to use, market depth 9999900.0 to 10000100.0, hurdle 0.504 percent; XRP buy 1.36152, sell 1.36658, 65 seconds old, okay to use, market depth 5708.7 to 5679.4, hurdle 0.875 percent; UNI buy 6.12552, sell 6.1384, 32 seconds old, okay to use, market depth 4350.1 to 4349.7, hurdle 0.713 percent; LTC buy 53.75, sell 54.0694, 110 seconds old, okay to use, market depth 3663.6 to 3654.2, hurdle 1.099 percent; HYPE buy 78.532, sell 78.915, 109 seconds old, okay to use, market depth 60606.3 to 60504.1, hurdle 0.992 percent; LINK buy 11.4841, sell 11.506, 116 seconds old, okay to use, market depth 13225.1 to 13362.2, hurdle 0.694 percent. There were 8 fresh quotes out of 10. The recipient had 598.50056519 BTC with a mempool net of 0.00014715 and 313 transactions. The reserve had 3601.47207156 BTC with a mempool net of 0.0 and 597 transactions. A recent page was 57, controlled 0, and did not have a full history. The tip was block 4053583, 87 seconds old, at time 1789183810. The second block was 4053584, 27 seconds old, at time 1789183870. There were 3 operator notes and an operator hash of 1e913157cc811e629b1b86d4db894412f924f07ea0928c7aa1a3e183681de0ca with a claim about a bitcoin theft from the Liquid Network. The fund attempted 7 out of 7 panels but found 0 eligible, as it was looking at balances, not sales or recovery.

A news scan found the headline 'Eight crypto hourly signals and missed entries through 06:15 (exit 0)'. The fund is called crypto and it is for research only, with 0 eligible entries. The observations were made on 2026-09-12T03:31:08.401408+00:00 and the source is https://runway.adityaperswal.workers.dev/data/crypto-bars. The following signals were observed: for Bitcoin (BTC) at 2026-09-12T02:00:00+00:00, the price was 77274.1215, the high 480 was 82233.502538071, and the signal was false. For Ethereum (ETH) at 2026-09-12T02:00:00+00:00, the price was 2514.923, the high 480 was 2664.8535, and the signal was false. For Solana (SOL) at 2026-09-12T02:00:00+00:00, the price was 101.8755, the high 480 was 110.5357, and the signal was false. For Litecoin (LTC) at 2026-09-12T02:00:00+00:00, the price was 53.99, the high 480 was 59.34355, and the signal was false. For XRP at 2026-09-12T02:00:00+00:00, the price was 1.362142495, the high 480 was 1.54995, and the signal was false. For HYPE at 2026-09-12T02:00:00+00:00, the price was 79.1, the high 480 was 89.9954713195, and the signal was false. For Chainlink (LINK) at 2026-09-12T02:00:00+00:00, the price was 11.51915, the high 480 was 13.677, and the signal was false. For Uniswap (UNI) at 2026-09-12T02:00:00+00:00, the price was 6.0967, the high 480 was 7.4795, and the signal was false. High 480 refers to the highest price in the last 480 minutes. The observations were made for research purposes only and no orders were placed.

A fund for crypto research was observed on 2026-09-12T03:16:11.752568+00:00. The source of the data is https://www.okx.com/api/v5/public/time with an offset of 0.0735 seconds and an uncertainty of 0.0939 seconds. Another source, https://www.okx.com/api/v5/public/funding-rate and https://www.okx.com/api/v5/public/open-interest, provided the following information: for Bitcoin, the funding rate was 0.357 basis points and the settled rate was 0.3247 basis points, with an open interest of 27153.344 coins. For Ethereum, the funding rate was 0.8082 basis points and the settled rate was 0.2743 basis points, with an open interest of 627499.194 coins. For Solana, the funding rate was 0.7595 basis points and the settled rate was 0.6055 basis points, with an open interest of 2741391.92 coins. A news scan found the headline from https://www.circle.com/transparency, which reported on September 10, 2026, that the issuer had 74.06 billion in circulation, with 6.66 billion issued and 6.91 billion redeemed in the last 7 days, 38.06 billion issued and 36.04 billion redeemed in the last 30 days, and 334.68 billion issued and 332.88 billion redeemed in the last 365 days. Another source, https://www.usdc.com/, could not be parsed due to ambiguity. A stablecoin tracker at https://stablecoins.llama.fi/stablecoin/2 reported that on September 10, 2026, there were 74,462,450,366.87 coins in circulation, with 46,937,295,105.4 on Ethereum, 7,249,020,091.29 on Solana, and 6,807,735,438.36 on Hyperliquid L1. The fund is not buying and is not eligible, with a status of 0.

A fund for crypto research was observed on 2026-09-12T03:01:54.065360+00:00. The data came from two sources: https://runway.adityaperswal.workers.dev/data/crypto-snapshots and https://blockstream.info/liquid/api/blocks. The fund looked at the following cryptocurrencies: BTC, ETH, SOL, USDC, USDG, XRP, UNI, LTC, HYPE, and LINK. For each one, it checked the bid and ask prices, the age of the data, whether the data was okay, the depth in US dollars, and a hurdle percentage. The results were: BTC bid 77225.6 and ask 77250.721 with age 52 and hurdle 0.535, ETH bid 2512.328 and ask 2512.8 with age 66 and hurdle 0.521, SOL bid 101.64 and ask 101.773 with age 52 and hurdle 0.633, USDC bid 0.9995 and ask 1.00001 with age 5305 and hurdle 0.553, USDG bid 0.99999 and ask 1.00001 with age 169792 and hurdle 0.504, XRP bid 1.36 and ask 1.363285 with age 52 and hurdle 0.745, UNI bid 6.07211 and ask 6.086 with age 51 and hurdle 0.732, LTC bid 53.74 and ask 54.065 with age 52 and hurdle 1.110, HYPE bid 78.64 and ask 79.03 with age 52 and hurdle 1.000, and LINK bid 11.49 and ask 11.51 with age 51 and hurdle 0.677. The fund got fresh quotes for 8 out of 10 of these cryptocurrencies. It also looked at the recipient and reserve balances for BTC, which were 598.50055859 and 3601.47207156 respectively. The fund checked the recent page, which was 58, and the tip, which was 4053554 with an age of 44 seconds. There was an error, a ConnectTimeout, when trying to connect to a second source. The operator notes had 3 entries and the operator hash was 1e913157cc811e629b1b86d4db894412f924f07ea0928c7aa1a3e183681de0ca. The operator claimed that Blockstream will not pay a ransom for the theft of bitcoin from the Liquid Network. The fund attempted to access 7 out of 7 panels and found 0 eligible, with balances but no sales or recovery.

A news scan found the following data line: this is a list of hourly signals for eight cryptocurrencies as of 06:15 on September 12, 2026. The fund is called crypto and it is for research only, meaning no actual trades are made, and it is not eligible, which means it is not being used to make investment decisions. The data comes from the website https://runway.adityaperswal.workers.dev/data/crypto-bars and was observed at 03:01:11 on September 12, 2026. Here are the details for each cryptocurrency: BTC had a price of 77274.1215, a high price over the last 480 minutes of 82233.502538071, and no signal to buy or sell, which means the conditions for a trade were not met. ETH had a price of 2514.923, a high price over the last 480 minutes of 2664.8535, and no signal, and it was available to trade recently, specifically at 14:00 on September 11, 2026. SOL had a price of 101.8755, a high price over the last 480 minutes of 110.5357, and no signal. LTC had a price of 53.99, a high price over the last 480 minutes of 59.34355, and no signal. XRP had a price of 1.362142495, a high price over the last 480 minutes of 1.54995, and no signal. HYPE had a price of 79.1, a high price over the last 480 minutes of 89.9954713195, and no signal. LINK had a price of 11.51915, a high price over the last 480 minutes of 13.677, and no signal. UNI had a price of 6.0967, a high price over the last 480 minutes of 7.4795, and no signal. This data is for observation only and no orders were placed.

A news scan found the following data: this is a research-only report, not a live asset, with 36 pairs of calendar-volume census data for USD inventory. The data source is https://runway.adityaperswal.workers.dev/data/crypto-bars, which includes 7 completed UTC days and ranks including stablecoins. The data for each cryptocurrency is as follows: USDC had a 7-day volume of $10,367,949, with 7 out of 7 days completed, a close price of 0.999755, and a prior 20-day average of 1.000005, with a signal of 0 and the last day not being zero. USDG had a 7-day volume of $1,899,981, with 2 out of 7 days completed, a close price of 1, and no prior 20-day average, with a signal of 0 and the last day not being zero. BTC had a 7-day volume of $677,156, with 7 out of 7 days completed, a close price of 77,212.832, and a prior 20-day average of 81,269.4915, with a signal of 0 and the last day not being zero. ETH had a 7-day volume of $459,605, with 7 out of 7 days completed, a close price of 2,516.066, and a prior 20-day average of 2,515.445, with a signal of 1 and the last day not being zero. LTC had a 7-day volume of $254,429, with 7 out of 7 days completed, a close price of 53.27, and a prior 20-day average of 55.20535, with a signal of 0 and the last day not being zero. XRP had a 7-day volume of $199,205, with 7 out of 7 days completed, a close price of 1.35664, and a prior 20-day average of 1.51966, with a signal of 0 and the last day not being zero. SOL had a 7-day volume of $192,379, with 7 out of 7 days completed, a close price of 102.4335, and a prior 20-day average of 109.15, with a signal of 0 and the last day not being zero. HYPE had a 7-day volume of $150,543, with 7 out of 7 days completed, a close price of 79.4395, and a prior 20-day average of 87.96325, with a signal of 0 and the last day not being zero. LINK had a 7-day volume of $146,046, with 7 out of 7 days completed, a close price of 11.5477, and a prior 20-day average of 13.2346, with a signal of 0 and the last day not being zero. UNI had a 7-day volume of $107,781, with 7 out of 7 days completed, a close price of 6.0076, and a prior 20-day average of 7.294845, with a signal of 0 and the last day not being zero.

A fund for crypto research made observations on September 12, 2026, at 02:31:49.659187 UTC. The data came from two sources: https://runway.adityaperswal.workers.dev/data/crypto-snapshots and https://blockstream.info/liquid/api/blocks. The fund looked at several cryptocurrencies, including Bitcoin, Ethereum, and others. For each one, it noted the current buy and sell prices, how old the data was, and whether it was okay to use. It also calculated a hurdle percentage, which is a measure of how hard it is to make a profit. The results were: BTC with a buy price of 77257.4 and a sell price of 77284.2, age of 43 seconds, and hurdle percentage of 0.537; ETH with a buy price of 2513.42 and a sell price of 2514.438, age of 481 seconds, and hurdle percentage of 0.543; SOL with a buy price of 101.76 and a sell price of 101.844, age of 396 seconds, and hurdle percentage of 0.585; USDC with a buy price of 0.9995 and a sell price of 1.00001, age of 3501 seconds, and hurdle percentage of 0.553; USDG with a buy price of 0.99999 and a sell price of 1.00001, age of 167988 seconds, and hurdle percentage of 0.504; XRP with a buy price of 1.35959 and a sell price of 1.363285, age of 76 seconds, and hurdle percentage of 0.775; UNI with a buy price of 6.0682 and a sell price of 6.07741, age of 67 seconds, and hurdle percentage of 0.654; LTC with a buy price of 53.76 and a sell price of 54.0984, age of 73 seconds, and hurdle percentage of 1.135; HYPE with a buy price of 78.8875 and a sell price of 79.2488, age of 100 seconds, and hurdle percentage of 0.962; and LINK with a buy price of 11.524 and a sell price of 11.55, age of 184 seconds, and hurdle percentage of 0.729. The fund got fresh quotes for 6 out of 10 cryptocurrencies. It also looked at the recipient and reserve balances for Bitcoin, which were 598.50055529 and 3601.47207156, respectively. The fund noted that there was an error connecting to a server, and it included a message from the operator about a theft of bitcoin from the Liquid Network. The fund did not make any sales or recover any balances.

A news scan found the headline 'job: Eight crypto hourly signals and missed entries through 06:15 (exit 0)'. The fund is called crypto and it is for research only, with 0 eligible entries. The observations were made on 2026-09-12T02:31:08.981580+00:00 from the source https://runway.adityaperswal.workers.dev/data/crypto-bars. The following signals were observed: for Bitcoin, the price was 77249.807 with a high of 82233.502538071, and no signal was given; for Ethereum, the price was 2514.15 with a high of 2664.8535, and no signal was given, with recent availability on 2026-09-11T14:00:00+00:00; for Solana, the price was 102.035 with a high of 110.5357, and no signal was given; for Litecoin, the price was 53.69345 with a high of 59.34355, and no signal was given; for XRP, the price was 1.360542495 with a high of 1.54995, and no signal was given; for HYPE, the price was 78.77 with a high of 89.9954713195, and no signal was given; for LINK, the price was 11.53 with a high of 13.677, and no signal was given; for UNI, the price was 6.048675 with a high of 7.4795, and no signal was given. The observations were for watching entry points only, with no orders made.

The decision, labeled as Decision2836, remains unchanged with no orders placed and a cash balance of $153.18 in crypto. A precision correction was made to the wording regarding ETH, which is currently priced at $2514.15 and needs a 5.9942128% increase to reach $2664.8535. This means it is 5.65523% below the threshold when the threshold is used as the denominator. The gap percentage fields are calculated using the formula (high/close - 1) * 100. The work was completed within 24 minutes of the 45-minute maximum allowed, and the final recording was reserved after 21 minutes with no overrun. A total of 192 diagnostic cells and all measured screen rows and CI records were completed. Lessons 812-814 covered technical, execution, and psyche topics. The next monitor is scheduled for 30904 UTC, with 234 local tasks passed and remote tasks pending. The costed comparators are not live edge due to entry-bar, full coverage, and 15-minute policy gaps. There is no new fund proposal. The report includes 36 pairs, 13,181 US listing names with 181 matches and 180 bars, as well as six non-US expressions. The primary exchanges are Circle, Metaplanet, and Hyperliquid, and Alpaca and OKX. It is recommended to choose cash for absent fresh entries and acknowledge the opportunity cost.

A test was done on the watch8.sh script by using the job229 script and making some changes. The changes included using the symbols BTC, ETH, SOL, LTC, XRP, HYPE, LINK, and UNI, and setting a cutoff date of September 12 at 06:15 UTC. The test also included a compact list of one-line symbols with information such as hour, close, high, and current signal. The test had the same completeness as before and a 24-hour missed signal loop, with a 12-page guard per symbol. The local test passed with 8 rows and an END signal. The job234 script was accepted after canceling job229, with a 15-minute timeout and 16 runs. The script is stored in the list_jobs234 for reuse, but does not include any orders or authenticated trading keys. An audit will be done on the remote END and a 30-minute observed cadence at 04 UTC on the 309th day. Jobs 229, 227, and 228 have been retired or retained. The full 3-year hour history needs more than 100 pages because the venue limit returns around 181-202 hour bars per page. A new fetch with a 200-page cap and a 429 backoff has resolved all 8 issues. The full historical list of jobs will not be read into the output, but rather filtered to show only active and selected scripts. The next steps include preserving the frozen ETH policy 2765 and creating a new comparator 2816-23, which is not a deployment validation. There is still an open issue with the ETH gap of 5.994% and the old missed path 2834. Other items to review include Arc 104/307, Metaplanet 308, net HYPE burn after issuance, and matched ETF share changes. The current cash is 153.18, not 235.86.

A news scan found the headline was not available, but a data line from a crypto research fund showed the following information. The fund, which is for crypto research only, was not eligible, and the observation was made on 2026-09-12 at 02:16:18.235316+00:00. The source of the information was https://www.okx.com/api/v5/public/time, which had an offset of 0.0584 seconds and an uncertainty of 0.1413 seconds. Another source, https://www.okx.com/api/v5/public/funding-rate and https://www.okx.com/api/v5/public/open-interest, provided the following data for three cryptocurrencies: BTC, ETH, and SOL. For BTC, the funding rate was 0.2798 basis points and the settled rate was 0.3247 basis points, with an open interest of 27225.147 coins. For ETH, the funding rate was 0.74 basis points and the settled rate was 0.2743 basis points, with an open interest of 626625.698 coins. For SOL, the funding rate was 0.8213 basis points and the settled rate was 0.6055 basis points, with an open interest of 2722688.59 coins. A news scan also found data from https://www.circle.com/transparency, which showed the issuer date was September 10, 2026, and the balance was 74.06, with issue and redeem amounts for 7, 30, and 365 days. Another source, https://www.usdc.com/, had a raw CMS value of 74342664805.879211266, but the issuer date was not parsed. A stablecoin data source, https://stablecoins.llama.fi/stablecoin/2, showed a circulating value of 74462450366.86803, with breakdowns for Ethereum, Solana, and Hyperliquid L1. The flow is not buying, and the eligible amount is 0.

A news scan found the text describing the C91202 preregistered static-exit comparator, which is not a replacement of the frozen ETH rule. Before results, the system ran tests on the top 10 most volatile assets, including BTC, ETH, LTC, XRP, SOL, HYPE, LINK, and UNI, using daily and hourly bars for the 15, 20, and 30 day high and half-lookback low. The system was trained on data from 2024 to 2025 and tested on a held-out time period from January 1, 2026, to September 11, 2026, which is a previously examined market that is not new, also known as out-of-sample or OOS. The next exact open entry was determined by the Average True Range, or ATR, of 14 days from completed prior days in Coordinated Universal Time, or UTC, with a stop loss set at 2 times the ATR and a target solved for a net 3 times the risk, or 3R, with a 20-day timeout. The stop and target were fixed, with no dynamic trailing stop that adds 2R. A 25 basis point fee was credited to each asset on each side, with 10 to 20 basis points of lippage, which refers to the difference between the expected price of a trade and the actual price. If the price gapped through the stop or target, the position was filled, with a dual hit stop triggered first. The system did not use live 15-minute checks for idealized stops within a bar, and did not allow same-bar entry and exit, which could lead to overly optimistic results. The system was explicitly a research comparator to quantify regimes, not a trade authorization system. All 24 cells and names were retained, with no selection of the best parameters. The vectorbt initialization cash and raw wealth assertion were less than 1e-9, which is a very small number. The historical top 10 selection survivorship remained, but HYPE lacked training data from 2024 to 2025. The frozen ETH policy and missed Friday path remained separate.

The C91202 question tree is being reviewed before searching. Initially, there was no new major entry and the edge for six trades in Ethereum was uncertain, so the desire to recover $349.01 was set aside. The resolution involved completing the daily and hourly bars in UTC time and timestamped quotes, and discarding partial bars and media return anecdotes. The question tree has four main parts: Can we execute what was tested, including checking the integrity of the universe, time, and capacity, with points such as published pairs versus live assets, missing or zero-volume days, completed hours, and missed signals, as well as current fund cap and bid/ask costs; Is there an independent edge, looking at momentum, leverage, and liquidity, including daily versus hourly definitions, funding historical percentile, and issuer issuance/redemptions periods; What structural source could change the conclusion, including adoption, security, and other expressions like US crypto equities/ETFs and non-US listings; Can the payoff survive being wrong, considering exits, sizing, and points like ATR and >=3R net, 15-minute gaps, half Kelly, and correlated positions.