[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"summary-941741f2e1ae4f3e-google-s-auto-diagnose-llm-diagnoses-test-failures-summary":3,"summaries-facets-categories":81,"summary-related-941741f2e1ae4f3e-google-s-auto-diagnose-llm-diagnoses-test-failures-summary":3666},{"id":4,"title":5,"ai":6,"body":13,"categories":46,"created_at":48,"date_modified":48,"description":40,"extension":49,"faq":48,"featured":50,"kicker_label":48,"meta":51,"navigation":63,"path":64,"published_at":65,"question":48,"scraped_at":66,"seo":67,"sitemap":68,"source_id":69,"source_name":70,"source_type":71,"source_url":72,"stem":73,"tags":74,"thumbnail_url":48,"tldr":78,"tweet":48,"unknown_tags":79,"__hash__":80},"summaries\u002Fsummaries\u002F941741f2e1ae4f3e-google-s-auto-diagnose-llm-diagnoses-test-failures-summary.md","Google's Auto-Diagnose: LLM Diagnoses Test Failures at 90% Accuracy",{"provider":7,"model":8,"input_tokens":9,"output_tokens":10,"processing_time_ms":11,"cost_usd":12},"openrouter","x-ai\u002Fgrok-4.1-fast",7775,1678,12431,0.00237135,{"type":14,"value":15,"toc":39},"minimark",[16,21,25,29,32,36],[17,18,20],"h2",{"id":19},"slash-integration-test-debug-time-with-llm-log-analysis","Slash Integration Test Debug Time with LLM Log Analysis",[22,23,24],"p",{},"Integration tests at Google, which are 78% functional per a 239-developer survey, often fail with generic symptoms like timeouts while root causes hide in SUT component logs amid noise. Developers report 38.4% of failures take over an hour to diagnose (vs. 2.7% for unit tests), and 8.9% exceed a day—top complaint in a 6,059-developer EngSat survey. Auto-Diagnose triggers on failure via pub\u002Fsub, aggregates INFO+ logs across data centers\u002Fprocesses\u002Fthreads, joins and sorts them by timestamp into one stream, adds component metadata, and feeds to Gemini 2.5 Flash (temperature=0.1, topp=0.8). This yields p50 latency of 56s and p90 of 346s, with 110k input\u002F6k output tokens per run, letting devs act before context-switching.",[17,26,28],{"id":27},"step-by-step-prompting-ensures-reliable-root-causes","Step-by-Step Prompting Ensures Reliable Root Causes",[22,30,31],{},"No fine-tuning needed—pure prompt engineering guides the LLM: scan sections, read context, locate failure, summarize errors, conclude only with evidence, and apply hard negatives like 'no conclusion if missing component logs.' Output post-processes to markdown with ==Conclusion== (root cause), ==Investigation Steps==, and ==Most Relevant Log Lines== (clickable links), auto-posted to Critique code reviews. Manual eval on 71 failures from 39 teams hit 90.14% root cause accuracy; failures exposed infra bugs like unsaved crash logs, fixed via feedback loop.",[17,33,35],{"id":34},"production-feedback-ranks-it-top-38-of-tools","Production Feedback Ranks It Top 3.8% of Tools",[22,37,38],{},"Deployed on 52,635 failing tests across 224,782 executions and 91,130 changes by 22,962 devs. Of 517 feedbacks from 437 devs, 84.3% were reviewer 'Please fix' requests; dev helpfulness 63%, 'Not helpful' just 5.8% (under 10% live threshold), #14 of 370 Critique tools (top 3.78%). Replicate by building similar pipelines: aggregate\u002Fsort logs, chain-of-thought prompt general LLMs, integrate with review tools for instant value.",{"title":40,"searchDepth":41,"depth":41,"links":42},"",2,[43,44,45],{"id":19,"depth":41,"text":20},{"id":27,"depth":41,"text":28},{"id":34,"depth":41,"text":35},[47],"AI & LLMs",null,"md",false,{"content_references":52,"triage":58},[53],{"type":54,"title":55,"url":56,"context":57},"paper","Auto-Diagnose Pre-Print","https:\u002F\u002Farxiv.org\u002Fpdf\u002F2604.12108","recommended",{"relevance":59,"novelty":60,"quality":60,"actionability":59,"composite":61,"reasoning":62},5,4,4.55,"Category: AI & LLMs. The article provides a detailed overview of Google's Auto-Diagnose system, which uses LLMs to improve integration test debugging, directly addressing the pain point of long diagnosis times for developers. It includes specific steps for replicating the system, making it highly actionable for the target audience.",true,"\u002Fsummaries\u002F941741f2e1ae4f3e-google-s-auto-diagnose-llm-diagnoses-test-failures-summary","2026-04-18 06:00:41","2026-04-19 01:22:38",{"title":5,"description":40},{"loc":64},"941741f2e1ae4f3e","MarkTechPost","article","https:\u002F\u002Fwww.marktechpost.com\u002F2026\u002F04\u002F17\u002Fgoogle-ai-releases-auto-diagnose-an-large-language-model-llm-based-system-to-diagnose-integration-test-failures-at-scale\u002F","summaries\u002F941741f2e1ae4f3e-google-s-auto-diagnose-llm-diagnoses-test-failures-summary",[75,76,77],"llm","prompt-engineering","dev-productivity","Prompt-engineer Gemini 2.5 Flash on timestamp-sorted logs to auto-diagnose integration test root causes, posting fixes to code reviews—90.14% accurate on 71 real failures, 5.8% 'Not helpful' in production across 52k+ tests.",[77],"5z_FYqr6PVfUMXcxaSSuLOcSDdpfdwoN6syFaCpzipE",[82,85,88,90,93,96,98,100,102,104,106,108,111,113,115,117,119,121,123,125,127,129,132,135,137,139,142,144,146,149,151,153,155,157,159,161,163,165,167,169,171,173,175,177,179,181,183,185,187,189,191,193,195,197,199,201,203,205,207,209,211,213,215,217,219,221,223,225,227,229,231,233,235,237,239,241,243,245,247,249,251,253,255,257,259,261,263,265,267,269,271,273,275,277,279,281,283,285,287,289,291,293,295,297,299,301,303,305,307,309,311,313,315,317,319,321,323,325,327,329,331,333,335,337,339,341,343,345,347,349,351,353,355,357,359,361,363,365,367,369,371,373,375,377,379,381,383,385,387,389,391,393,395,397,399,401,403,406,408,410,412,414,416,418,420,422,424,426,428,430,432,434,436,438,440,442,444,446,448,450,452,454,456,458,460,462,464,466,468,470,472,474,476,478,480,482,484,486,488,490,492,494,496,498,500,502,504,506,508,510,512,514,516,518,520,522,524,526,528,530,532,534,536,538,540,542,544,546,548,550,552,554,556,558,560,562,564,566,568,570,572,574,576,578,580,582,584,586,588,590,592,594,596,598,600,602,604,606,608,610,612,614,616,618,620,622,624,626,628,630,632,634,636,638,640,642,644,646,648,650,652,654,656,658,660,662,664,666,668,670,672,674,676,678,680,682,684,686,688,690,692,694,696,698,700,702,704,706,708,710,712,714,716,718,720,722,724,726,728,730,732,734,736,738,740,742,744,746,748,750,752,754,756,758,760,762,764,766,768,770,772,774,776,778,780,782,784,786,788,790,792,794,796,798,800,802,804,806,808,810,812,814,816,818,820,822,824,826,828,830,832,834,836,838,840,842,844,846,848,850,852,854,856,858,860,862,864,866,868,870,872,874,876,878,880,882,884,886,888,890,892,894,896,898,900,902,904,906,908,910,912,914,916,918,920,922,924,926,928,930,932,934,936,938,940,942,944,946,948,950,952,954,956,958,960,962,964,966,968,970,972,974,976,978,980,982,984,986,988,990,992,994,996,998,1000,1002,1004,1006,1008,1010,1012,1014,1016,1018,1020,1022,1024,1026,1028,1030,1032,1034,1036,1038,1040,1042,1044,1046,1048,1050,1052,1054,1056,1058,1060,1062,1064,1066,1068,1070,1072,1074,1076,1078,1080,1082,1084,1086,1088,1090,1092,1094,1096,1098,1100,1102,1104,1106,1108,1110,1112,1114,1116,1118,1120,1122,1124,1126,1128,1130,1132,1134,1136,1138,1140,1142,1144,1146,1148,1150,1152,1154,1156,1158,1160,1162,1164,1166,1168,1170,1172,1174,1176,1178,1180,1182,1184,1186,1188,1190,1192,1194,1196,1198,1200,1202,1204,1206,1208,1210,1212,1214,1216,1218,1220,1222,1224,1226,1228,1230,1232,1234,1236,1238,1240,1242,1244,1246,1248,1250,1252,1254,1256,1258,1260,1262,1264,1266,1268,1270,1272,1274,1276,1278,1280,1282,1284,1286,1288,1290,1292,1294,1296,1298,1300,1302,1304,1306,1308,1310,1312,1314,1316,1318,1320,1322,1324,1326,1328,1330,1332,1334,1336,1338,1340,1342,1344,1346,1348,1350,1352,1354,1356,1358,1360,1362,1364,1366,1368,1370,1372,1374,1376,1378,1380,1382,1384,1386,1388,1390,1392,1394,1396,1398,1400,1402,1404,1406,1408,1410,1412,1414,1416,1418,1420,1422,1424,1426,1428,1430,1432,1434,1436,1438,1440,1442,1444,1446,1448,1450,1452,1454,1456,1458,1460,1462,1464,1466,1468,1470,1472,1474,1476,1478,1480,1482,1484,1486,1488,1490,1492,1494,1496,1498,1500,1502,1504,1506,1508,1510,1512,1514,1516,1518,1520,1522,1524,1526,1528,1530,1532,1534,1536,1538,1540,1542,1544,1546,1548,1550,1552,1554,1556,1558,1560,1562,1564,1566,1568,1570,1572,1574,1576,1578,1580,1582,1584,1586,1588,1590,1592,1594,1596,1598,1600,1602,1604,1606,1608,1610,1612,1614,1616,1618,1620,1622,1624,1626,1628,1630,1632,1634,1636,1638,1640,1642,1644,1646,1648,1650,1652,1654,1656,1658,1660,1662,1664,1666,1668,1670,1672,1674,1676,1678,1680,1682,1684,1686,1688,1690,1692,1694,1696,1698,1700,1702,1704,1706,1708,1710,1712,1714,1716,1718,1720,1722,1724,1726,1728,1730,1732,1734,1736,1738,1740,1742,1744,1746,1748,1750,1752,1754,1756,1758,1760,1762,1764,1766,1768,1770,1772,1774,1776,1778,1780,1782,1784,1786,1788,1790,1792,1794,1796,1798,1800,1802,1804,1806,1808,1810,1812,1814,1816,1818,1820,1822,1824,1826,1828,1830,1832,1834,1836,1838,1840,1842,1844,1846,1848,1850,1852,1854,1856,1858,1860,1862,1864,1866,1868,1870,1872,1874,1876,1878,1880,1882,1884,1886,1888,1890,1892,1894,1896,1898,1900,1902,1904,1906,1908,1910,1912,1914,1916,1918,1920,1922,1924,1926,1928,1930,1932,1934,1936,1938,1940,1942,1944,1946,1948,1950,1952,1954,1956,1958,1960,1962,1964,1966,1968,1970,1972,1974,1976,1978,1980,1982,1984,1986,1988,1990,1992,1994,1996,1998,2000,2002,2004,2006,2008,2010,2012,2014,2016,2018,2020,2022,2024,2026,2028,2030,2032,2034,2036,2038,2040,2042,2044,2046,2048,2050,2052,2054,2056,2058,2060,2062,2064,2066,2068,2070,2072,2074,2076,2078,2080,2082,2084,2086,2088,2090,2092,2094,2096,2098,2100,2102,2104,2106,2108,2110,2112,2114,2116,2118,2120,2122,2124,2126,2128,2130,2132,2134,2136,2138,2140,2142,2144,2146,2148,2150,2152,2154,2156,2158,2160,2162,2164,2166,2168,2170,2172,2174,2176,2178,2180,2182,2184,2186,2188,2190,2192,2194,2196,2198,2200,2202,2204,2206,2208,2210,2212,2214,2216,2218,2220,2222,2224,2226,2228,2230,2232,2234,2236,2238,2240,2242,2244,2246,2248,2250,2252,2254,2256,2258,2260,2262,2264,2266,2268,2270,2272,2274,2276,2278,2280,2282,2284,2286,2288,2290,2292,2294,2296,2298,2300,2302,2304,2306,2308,2310,2312,2314,2316,2318,2320,2322,2324,2326,2328,2330,2332,2334,2336,2338,2340,2342,2344,2346,2348,2350,2352,2354,2356,2358,2360,2362,2364,2366,2368,2370,2372,2374,2376,2378,2380,2382,2384,2386,2388,2390,2392,2394,2396,2398,2400,2402,2404,2406,2408,2410,2412,2414,2416,2418,2420,2422,2424,2426,2428,2430,2432,2434,2436,2438,2440,2442,2444,2446,2448,2450,2452,2454,2456,2458,2460,2462,2464,2466,2468,2470,2472,2474,2476,2478,2480,2482,2484,2486,2488,2490,2492,2494,2496,2498,2500,2502,2504,2506,2508,2510,2512,2514,2516,2518,2520,2522,2524,2526,2528,2530,2532,2534,2536,2538,2540,2542,2544,2546,2548,2550,2552,2554,2556,2558,2560,2562,2564,2566,2568,2570,2572,2574,2576,2578,2580,2582,2584,2586,2588,2590,2592,2594,2596,2598,2600,2602,2604,2606,2608,2610,2612,2614,2616,2618,2620,2622,2624,2626,2628,2630,2632,2634,2636,2638,2640,2642,2644,2646,2648,2650,2652,2654,2656,2658,2660,2662,2664,2666,2668,2670,2672,2674,2676,2678,2680,2682,2684,2686,2688,2690,2692,2694,2696,2698,2700,2702,2704,2706,2708,2710,2712,2714,2716,2718,2720,2722,2724,2726,2728,2730,2732,2734,2736,2738,2740,2742,2744,2746,2748,2750,2752,2754,2756,2758,2760,2762,2764,2766,2768,2770,2772,2774,2776,2778,2780,2782,2784,2786,2788,2790,2792,2794,2796,2798,2800,2802,2804,2806,2808,2810,2812,2814,2816,2818,2820,2822,2824,2826,2828,2830,2832,2834,2836,2838,2840,2842,2844,2846,2848,2850,2852,2854,2856,2858,2860,2862,2864,2866,2868,2870,2872,2874,2876,2878,2880,2882,2884,2886,2888,2890,2892,2894,2896,2898,2900,2902,2904,2906,2908,2910,2912,2914,2916,2918,2920,2922,2924,2926,2928,2930,2932,2934,2936,2938,2940,2942,2944,2946,2948,2950,2952,2954,2956,2958,2960,2962,2964,2966,2968,2970,2972,2974,2976,2978,2980,2982,2984,2986,2988,2990,2992,2994,2996,2998,3000,3002,3004,3006,3008,3010,3012,3014,3016,3018,3020,3022,3024,3026,3028,3030,3032,3034,3036,3038,3040,3042,3044,3046,3048,3050,3052,3054,3056,3058,3060,3062,3064,3066,3068,3070,3072,3074,3076,3078,3080,3082,3084,3086,3088,3090,3092,3094,3096,3098,3100,3102,3104,3106,3108,3110,3112,3114,3116,3118,3120,3122,3124,3126,3128,3130,3132,3134,3136,3138,3140,3142,3144,3146,3148,3150,3152,3154,3156,3158,3160,3162,3164,3166,3168,3170,3172,3174,3176,3178,3180,3182,3184,3186,3188,3190,3192,3194,3196,3198,3200,3202,3204,3206,3208,3210,3212,3214,3216,3218,3220,3222,3224,3226,3228,3230,3232,3234,3236,3238,3240,3242,3244,3246,3248,3250,3252,3254,3256,3258,3260,3262,3264,3266,3268,3270,3272,3274,3276,3278,3280,3282,3284,3286,3288,3290,3292,3294,3296,3298,3300,3302,3304,3306,3308,3310,3312,3314,3316,3318,3320,3322,3324,3326,3328,3330,3332,3334,3336,3338,3340,3342,3344,3346,3348,3350,3352,3354,3356,3358,3360,3362,3364,3366,3368,3370,3372,3374,3376,3378,3380,3382,3384,3386,3388,3390,3392,3394,3396,3398,3400,3402,3404,3406,3408,3410,3412,3414,3416,3418,3420,3422,3424,3426,3428,3430,3432,3434,3436,3438,3440,3442,3444,3446,3448,3450,3452,3454,3456,3458,3460,3462,3464,3466,3468,3470,3472,3474,3476,3478,3480,3482,3484,3486,3488,3490,3492,3494,3496,3498,3500,3502,3504,3506,3508,3510,3512,3514,3516,3518,3520,3522,3524,3526,3528,3530,3532,3534,3536,3538,3540,3542,3544,3546,3548,3550,3552,3554,3556,3558,3560,3562,3564,3566,3568,3570,3572,3574,3576,3578,3580,3582,3584,3586,3588,3590,3592,3594,3596,3598,3600,3602,3604,3606,3608,3610,3612,3614,3616,3618,3620,3622,3624,3626,3628,3630,3632,3634,3636,3638,3640,3642,3644,3646,3648,3650,3652,3654,3656,3658,3660,3662,3664],{"categories":83},[84],"Developer Productivity",{"categories":86},[87],"Business & SaaS",{"categories":89},[47],{"categories":91},[92],"AI Automation",{"categories":94},[95],"Product Strategy",{"categories":97},[47],{"categories":99},[84],{"categories":101},[87],{"categories":103},[],{"categories":105},[47],{"categories":107},[],{"categories":109},[110],"AI News & Trends",{"categories":112},[92],{"categories":114},[110],{"categories":116},[92],{"categories":118},[92],{"categories":120},[47],{"categories":122},[47],{"categories":124},[110],{"categories":126},[47],{"categories":128},[],{"categories":130},[131],"Design & Frontend",{"categories":133},[134],"Data Science & Visualization",{"categories":136},[110],{"categories":138},[],{"categories":140},[141],"Software Engineering",{"categories":143},[47],{"categories":145},[92],{"categories":147},[148],"Marketing & Growth",{"categories":150},[47],{"categories":152},[92],{"categories":154},[],{"categories":156},[],{"categories":158},[131],{"categories":160},[92],{"categories":162},[84],{"categories":164},[131],{"categories":166},[47],{"categories":168},[92],{"categories":170},[110],{"categories":172},[],{"categories":174},[],{"categories":176},[92],{"categories":178},[141],{"categories":180},[],{"categories":182},[87],{"categories":184},[],{"categories":186},[],{"categories":188},[92],{"categories":190},[92],{"categories":192},[47],{"categories":194},[],{"categories":196},[141],{"categories":198},[],{"categories":200},[],{"categories":202},[],{"categories":204},[47],{"categories":206},[148],{"categories":208},[131],{"categories":210},[131],{"categories":212},[47],{"categories":214},[92],{"categories":216},[47],{"categories":218},[47],{"categories":220},[92],{"categories":222},[92],{"categories":224},[134],{"categories":226},[110],{"categories":228},[92],{"categories":230},[148],{"categories":232},[92],{"categories":234},[95],{"categories":236},[],{"categories":238},[92],{"categories":240},[],{"categories":242},[92],{"categories":244},[141],{"categories":246},[131],{"categories":248},[47],{"categories":250},[],{"categories":252},[],{"categories":254},[92],{"categories":256},[],{"categories":258},[47],{"categories":260},[],{"categories":262},[84],{"categories":264},[141],{"categories":266},[87],{"categories":268},[110],{"categories":270},[47],{"categories":272},[],{"categories":274},[47],{"categories":276},[],{"categories":278},[141],{"categories":280},[134],{"categories":282},[],{"categories":284},[47],{"categories":286},[131],{"categories":288},[],{"categories":290},[131],{"categories":292},[92],{"categories":294},[],{"categories":296},[92],{"categories":298},[110],{"categories":300},[87],{"categories":302},[47],{"categories":304},[],{"categories":306},[92],{"categories":308},[47],{"categories":310},[95],{"categories":312},[],{"categories":314},[47],{"categories":316},[92],{"categories":318},[92],{"categories":320},[],{"categories":322},[134],{"categories":324},[47],{"categories":326},[],{"categories":328},[84],{"categories":330},[87],{"categories":332},[47],{"categories":334},[92],{"categories":336},[141],{"categories":338},[47],{"categories":340},[],{"categories":342},[],{"categories":344},[47],{"categories":346},[],{"categories":348},[131],{"categories":350},[],{"categories":352},[47],{"categories":354},[],{"categories":356},[92],{"categories":358},[47],{"categories":360},[131],{"categories":362},[],{"categories":364},[47],{"categories":366},[47],{"categories":368},[87],{"categories":370},[92],{"categories":372},[47],{"categories":374},[131],{"categories":376},[92],{"categories":378},[],{"categories":380},[],{"categories":382},[110],{"categories":384},[],{"categories":386},[47],{"categories":388},[87,148],{"categories":390},[],{"categories":392},[47],{"categories":394},[],{"categories":396},[],{"categories":398},[47],{"categories":400},[],{"categories":402},[47],{"categories":404},[405],"DevOps & Cloud",{"categories":407},[],{"categories":409},[110],{"categories":411},[131],{"categories":413},[],{"categories":415},[110],{"categories":417},[110],{"categories":419},[47],{"categories":421},[148],{"categories":423},[],{"categories":425},[87],{"categories":427},[],{"categories":429},[47,405],{"categories":431},[47],{"categories":433},[47],{"categories":435},[92],{"categories":437},[47,141],{"categories":439},[134],{"categories":441},[47],{"categories":443},[148],{"categories":445},[92],{"categories":447},[92],{"categories":449},[],{"categories":451},[92],{"categories":453},[47,87],{"categories":455},[],{"categories":457},[131],{"categories":459},[131],{"categories":461},[],{"categories":463},[],{"categories":465},[110],{"categories":467},[],{"categories":469},[84],{"categories":471},[141],{"categories":473},[47],{"categories":475},[131],{"categories":477},[92],{"categories":479},[141],{"categories":481},[110],{"categories":483},[131],{"categories":485},[],{"categories":487},[47],{"categories":489},[47],{"categories":491},[47],{"categories":493},[110],{"categories":495},[84],{"categories":497},[47],{"categories":499},[92],{"categories":501},[405],{"categories":503},[131],{"categories":505},[92],{"categories":507},[],{"categories":509},[],{"categories":511},[131],{"categories":513},[110],{"categories":515},[134],{"categories":517},[],{"categories":519},[47],{"categories":521},[47],{"categories":523},[87],{"categories":525},[47],{"categories":527},[47],{"categories":529},[110],{"categories":531},[],{"categories":533},[92],{"categories":535},[141],{"categories":537},[],{"categories":539},[47],{"categories":541},[47],{"categories":543},[92],{"categories":545},[],{"categories":547},[],{"categories":549},[47],{"categories":551},[],{"categories":553},[87],{"categories":555},[92],{"categories":557},[],{"categories":559},[84],{"categories":561},[47],{"categories":563},[87],{"categories":565},[110],{"categories":567},[],{"categories":569},[],{"categories":571},[],{"categories":573},[110],{"categories":575},[110],{"categories":577},[],{"categories":579},[],{"categories":581},[87],{"categories":583},[],{"categories":585},[],{"categories":587},[84],{"categories":589},[],{"categories":591},[148],{"categories":593},[92],{"categories":595},[87],{"categories":597},[92],{"categories":599},[141],{"categories":601},[],{"categories":603},[95],{"categories":605},[131],{"categories":607},[141],{"categories":609},[47],{"categories":611},[92],{"categories":613},[87],{"categories":615},[47],{"categories":617},[],{"categories":619},[],{"categories":621},[141],{"categories":623},[134],{"categories":625},[95],{"categories":627},[92],{"categories":629},[47],{"categories":631},[],{"categories":633},[405],{"categories":635},[],{"categories":637},[92],{"categories":639},[],{"categories":641},[],{"categories":643},[47],{"categories":645},[131],{"categories":647},[148],{"categories":649},[92],{"categories":651},[],{"categories":653},[84],{"categories":655},[],{"categories":657},[110],{"categories":659},[47,405],{"categories":661},[110],{"categories":663},[47],{"categories":665},[87],{"categories":667},[47],{"categories":669},[],{"categories":671},[87],{"categories":673},[],{"categories":675},[141],{"categories":677},[131],{"categories":679},[110],{"categories":681},[134],{"categories":683},[84],{"categories":685},[47],{"categories":687},[141],{"categories":689},[],{"categories":691},[],{"categories":693},[95],{"categories":695},[],{"categories":697},[47],{"categories":699},[],{"categories":701},[131],{"categories":703},[131],{"categories":705},[131],{"categories":707},[],{"categories":709},[],{"categories":711},[110],{"categories":713},[92],{"categories":715},[47],{"categories":717},[47],{"categories":719},[47],{"categories":721},[87],{"categories":723},[47],{"categories":725},[],{"categories":727},[141],{"categories":729},[141],{"categories":731},[87],{"categories":733},[],{"categories":735},[47],{"categories":737},[47],{"categories":739},[87],{"categories":741},[110],{"categories":743},[148],{"categories":745},[92],{"categories":747},[],{"categories":749},[131],{"categories":751},[],{"categories":753},[47],{"categories":755},[],{"categories":757},[87],{"categories":759},[92],{"categories":761},[],{"categories":763},[405],{"categories":765},[134],{"categories":767},[141],{"categories":769},[148],{"categories":771},[141],{"categories":773},[92],{"categories":775},[],{"categories":777},[],{"categories":779},[92],{"categories":781},[84],{"categories":783},[92],{"categories":785},[95],{"categories":787},[87],{"categories":789},[],{"categories":791},[47],{"categories":793},[95],{"categories":795},[47],{"categories":797},[47],{"categories":799},[148],{"categories":801},[131],{"categories":803},[92],{"categories":805},[],{"categories":807},[],{"categories":809},[405],{"categories":811},[141],{"categories":813},[],{"categories":815},[92],{"categories":817},[47],{"categories":819},[131,47],{"categories":821},[84],{"categories":823},[],{"categories":825},[47],{"categories":827},[84],{"categories":829},[131],{"categories":831},[92],{"categories":833},[141],{"categories":835},[],{"categories":837},[47],{"categories":839},[],{"categories":841},[84],{"categories":843},[],{"categories":845},[92],{"categories":847},[95],{"categories":849},[47],{"categories":851},[47],{"categories":853},[131],{"categories":855},[92],{"categories":857},[405],{"categories":859},[131],{"categories":861},[92],{"categories":863},[47],{"categories":865},[47],{"categories":867},[47],{"categories":869},[110],{"categories":871},[],{"categories":873},[95],{"categories":875},[92],{"categories":877},[131],{"categories":879},[92],{"categories":881},[141],{"categories":883},[131],{"categories":885},[92],{"categories":887},[110],{"categories":889},[],{"categories":891},[47],{"categories":893},[131],{"categories":895},[47],{"categories":897},[84],{"categories":899},[110],{"categories":901},[47],{"categories":903},[148],{"categories":905},[47],{"categories":907},[47],{"categories":909},[92],{"categories":911},[92],{"categories":913},[47],{"categories":915},[92],{"categories":917},[131],{"categories":919},[47],{"categories":921},[],{"categories":923},[],{"categories":925},[141],{"categories":927},[],{"categories":929},[84],{"categories":931},[405],{"categories":933},[],{"categories":935},[84],{"categories":937},[87],{"categories":939},[148],{"categories":941},[],{"categories":943},[87],{"categories":945},[],{"categories":947},[],{"categories":949},[],{"categories":951},[],{"categories":953},[],{"categories":955},[47],{"categories":957},[92],{"categories":959},[405],{"categories":961},[84],{"categories":963},[47],{"categories":965},[141],{"categories":967},[95],{"categories":969},[47],{"categories":971},[148],{"categories":973},[47],{"categories":975},[47],{"categories":977},[47],{"categories":979},[47,84],{"categories":981},[141],{"categories":983},[141],{"categories":985},[131],{"categories":987},[47],{"categories":989},[],{"categories":991},[],{"categories":993},[],{"categories":995},[141],{"categories":997},[134],{"categories":999},[110],{"categories":1001},[131],{"categories":1003},[],{"categories":1005},[47],{"categories":1007},[47],{"categories":1009},[],{"categories":1011},[],{"categories":1013},[92],{"categories":1015},[47],{"categories":1017},[87],{"categories":1019},[],{"categories":1021},[84],{"categories":1023},[47],{"categories":1025},[84],{"categories":1027},[47],{"categories":1029},[141],{"categories":1031},[148],{"categories":1033},[47,131],{"categories":1035},[110],{"categories":1037},[131],{"categories":1039},[],{"categories":1041},[405],{"categories":1043},[131],{"categories":1045},[92],{"categories":1047},[],{"categories":1049},[],{"categories":1051},[],{"categories":1053},[],{"categories":1055},[141],{"categories":1057},[92],{"categories":1059},[92],{"categories":1061},[405],{"categories":1063},[47],{"categories":1065},[47],{"categories":1067},[47],{"categories":1069},[],{"categories":1071},[131],{"categories":1073},[],{"categories":1075},[],{"categories":1077},[92],{"categories":1079},[],{"categories":1081},[],{"categories":1083},[148],{"categories":1085},[148],{"categories":1087},[92],{"categories":1089},[],{"categories":1091},[47],{"categories":1093},[47],{"categories":1095},[141],{"categories":1097},[131],{"categories":1099},[131],{"categories":1101},[92],{"categories":1103},[84],{"categories":1105},[47],{"categories":1107},[131],{"categories":1109},[131],{"categories":1111},[92],{"categories":1113},[92],{"categories":1115},[47],{"categories":1117},[],{"categories":1119},[],{"categories":1121},[47],{"categories":1123},[92],{"categories":1125},[110],{"categories":1127},[141],{"categories":1129},[84],{"categories":1131},[47],{"categories":1133},[],{"categories":1135},[92],{"categories":1137},[92],{"categories":1139},[],{"categories":1141},[84],{"categories":1143},[47],{"categories":1145},[84],{"categories":1147},[84],{"categories":1149},[],{"categories":1151},[],{"categories":1153},[92],{"categories":1155},[92],{"categories":1157},[47],{"categories":1159},[47],{"categories":1161},[110],{"categories":1163},[134],{"categories":1165},[95],{"categories":1167},[110],{"categories":1169},[131],{"categories":1171},[],{"categories":1173},[110],{"categories":1175},[],{"categories":1177},[],{"categories":1179},[],{"categories":1181},[],{"categories":1183},[141],{"categories":1185},[134],{"categories":1187},[],{"categories":1189},[47],{"categories":1191},[47],{"categories":1193},[134],{"categories":1195},[141],{"categories":1197},[],{"categories":1199},[],{"categories":1201},[92],{"categories":1203},[110],{"categories":1205},[110],{"categories":1207},[92],{"categories":1209},[84],{"categories":1211},[47,405],{"categories":1213},[],{"categories":1215},[131],{"categories":1217},[84],{"categories":1219},[92],{"categories":1221},[131],{"categories":1223},[],{"categories":1225},[92],{"categories":1227},[92],{"categories":1229},[47],{"categories":1231},[148],{"categories":1233},[141],{"categories":1235},[131],{"categories":1237},[],{"categories":1239},[92],{"categories":1241},[47],{"categories":1243},[92],{"categories":1245},[92],{"categories":1247},[92],{"categories":1249},[148],{"categories":1251},[92],{"categories":1253},[47],{"categories":1255},[],{"categories":1257},[148],{"categories":1259},[110],{"categories":1261},[92],{"categories":1263},[],{"categories":1265},[],{"categories":1267},[47],{"categories":1269},[92],{"categories":1271},[110],{"categories":1273},[92],{"categories":1275},[],{"categories":1277},[],{"categories":1279},[],{"categories":1281},[92],{"categories":1283},[],{"categories":1285},[],{"categories":1287},[134],{"categories":1289},[47],{"categories":1291},[134],{"categories":1293},[110],{"categories":1295},[47],{"categories":1297},[47],{"categories":1299},[92],{"categories":1301},[47],{"categories":1303},[],{"categories":1305},[],{"categories":1307},[405],{"categories":1309},[],{"categories":1311},[],{"categories":1313},[84],{"categories":1315},[],{"categories":1317},[],{"categories":1319},[],{"categories":1321},[],{"categories":1323},[141],{"categories":1325},[110],{"categories":1327},[148],{"categories":1329},[87],{"categories":1331},[47],{"categories":1333},[47],{"categories":1335},[87],{"categories":1337},[],{"categories":1339},[131],{"categories":1341},[92],{"categories":1343},[87],{"categories":1345},[47],{"categories":1347},[47],{"categories":1349},[84],{"categories":1351},[],{"categories":1353},[84],{"categories":1355},[47],{"categories":1357},[148],{"categories":1359},[92],{"categories":1361},[110],{"categories":1363},[87],{"categories":1365},[47],{"categories":1367},[92],{"categories":1369},[],{"categories":1371},[47],{"categories":1373},[84],{"categories":1375},[47],{"categories":1377},[],{"categories":1379},[110],{"categories":1381},[47],{"categories":1383},[],{"categories":1385},[87],{"categories":1387},[47],{"categories":1389},[],{"categories":1391},[],{"categories":1393},[],{"categories":1395},[47],{"categories":1397},[],{"categories":1399},[405],{"categories":1401},[47],{"categories":1403},[],{"categories":1405},[47],{"categories":1407},[47],{"categories":1409},[47],{"categories":1411},[47,405],{"categories":1413},[47],{"categories":1415},[47],{"categories":1417},[131],{"categories":1419},[92],{"categories":1421},[],{"categories":1423},[92],{"categories":1425},[47],{"categories":1427},[47],{"categories":1429},[47],{"categories":1431},[84],{"categories":1433},[84],{"categories":1435},[141],{"categories":1437},[131],{"categories":1439},[92],{"categories":1441},[],{"categories":1443},[47],{"categories":1445},[110],{"categories":1447},[47],{"categories":1449},[87],{"categories":1451},[],{"categories":1453},[405],{"categories":1455},[131],{"categories":1457},[131],{"categories":1459},[92],{"categories":1461},[110],{"categories":1463},[92],{"categories":1465},[47],{"categories":1467},[],{"categories":1469},[47],{"categories":1471},[],{"categories":1473},[],{"categories":1475},[47],{"categories":1477},[47],{"categories":1479},[47],{"categories":1481},[92],{"categories":1483},[47],{"categories":1485},[],{"categories":1487},[134],{"categories":1489},[92],{"categories":1491},[],{"categories":1493},[],{"categories":1495},[47],{"categories":1497},[110],{"categories":1499},[],{"categories":1501},[131],{"categories":1503},[405],{"categories":1505},[110],{"categories":1507},[141],{"categories":1509},[141],{"categories":1511},[110],{"categories":1513},[110],{"categories":1515},[405],{"categories":1517},[],{"categories":1519},[110],{"categories":1521},[47],{"categories":1523},[84],{"categories":1525},[110],{"categories":1527},[],{"categories":1529},[134],{"categories":1531},[110],{"categories":1533},[141],{"categories":1535},[110],{"categories":1537},[405],{"categories":1539},[47],{"categories":1541},[47],{"categories":1543},[],{"categories":1545},[87],{"categories":1547},[],{"categories":1549},[],{"categories":1551},[47],{"categories":1553},[47],{"categories":1555},[47],{"categories":1557},[47],{"categories":1559},[],{"categories":1561},[134],{"categories":1563},[84],{"categories":1565},[],{"categories":1567},[47],{"categories":1569},[47],{"categories":1571},[405],{"categories":1573},[405],{"categories":1575},[],{"categories":1577},[92],{"categories":1579},[110],{"categories":1581},[110],{"categories":1583},[47],{"categories":1585},[92],{"categories":1587},[],{"categories":1589},[131],{"categories":1591},[47],{"categories":1593},[47],{"categories":1595},[],{"categories":1597},[],{"categories":1599},[405],{"categories":1601},[47],{"categories":1603},[141],{"categories":1605},[87],{"categories":1607},[47],{"categories":1609},[],{"categories":1611},[92],{"categories":1613},[84],{"categories":1615},[84],{"categories":1617},[],{"categories":1619},[47],{"categories":1621},[131],{"categories":1623},[92],{"categories":1625},[],{"categories":1627},[47],{"categories":1629},[47],{"categories":1631},[92],{"categories":1633},[],{"categories":1635},[92],{"categories":1637},[141],{"categories":1639},[],{"categories":1641},[47],{"categories":1643},[],{"categories":1645},[47],{"categories":1647},[],{"categories":1649},[47],{"categories":1651},[47],{"categories":1653},[],{"categories":1655},[47],{"categories":1657},[110],{"categories":1659},[47],{"categories":1661},[47],{"categories":1663},[84],{"categories":1665},[47],{"categories":1667},[110],{"categories":1669},[92],{"categories":1671},[],{"categories":1673},[47],{"categories":1675},[148],{"categories":1677},[],{"categories":1679},[],{"categories":1681},[],{"categories":1683},[84],{"categories":1685},[110],{"categories":1687},[92],{"categories":1689},[47],{"categories":1691},[131],{"categories":1693},[92],{"categories":1695},[],{"categories":1697},[92],{"categories":1699},[],{"categories":1701},[47],{"categories":1703},[92],{"categories":1705},[47],{"categories":1707},[],{"categories":1709},[47],{"categories":1711},[47],{"categories":1713},[110],{"categories":1715},[131],{"categories":1717},[92],{"categories":1719},[131],{"categories":1721},[87],{"categories":1723},[],{"categories":1725},[],{"categories":1727},[47],{"categories":1729},[84],{"categories":1731},[110],{"categories":1733},[],{"categories":1735},[],{"categories":1737},[141],{"categories":1739},[131],{"categories":1741},[],{"categories":1743},[47],{"categories":1745},[],{"categories":1747},[148],{"categories":1749},[47],{"categories":1751},[405],{"categories":1753},[141],{"categories":1755},[],{"categories":1757},[92],{"categories":1759},[47],{"categories":1761},[92],{"categories":1763},[92],{"categories":1765},[47],{"categories":1767},[],{"categories":1769},[84],{"categories":1771},[47],{"categories":1773},[87],{"categories":1775},[141],{"categories":1777},[131],{"categories":1779},[],{"categories":1781},[],{"categories":1783},[],{"categories":1785},[92],{"categories":1787},[131],{"categories":1789},[110],{"categories":1791},[47],{"categories":1793},[110],{"categories":1795},[131],{"categories":1797},[],{"categories":1799},[131],{"categories":1801},[110],{"categories":1803},[87],{"categories":1805},[47],{"categories":1807},[110],{"categories":1809},[148],{"categories":1811},[],{"categories":1813},[],{"categories":1815},[134],{"categories":1817},[47,141],{"categories":1819},[110],{"categories":1821},[47],{"categories":1823},[92],{"categories":1825},[92],{"categories":1827},[47],{"categories":1829},[],{"categories":1831},[141],{"categories":1833},[47],{"categories":1835},[134],{"categories":1837},[92],{"categories":1839},[148],{"categories":1841},[405],{"categories":1843},[],{"categories":1845},[84],{"categories":1847},[92],{"categories":1849},[92],{"categories":1851},[141],{"categories":1853},[47],{"categories":1855},[47],{"categories":1857},[],{"categories":1859},[],{"categories":1861},[],{"categories":1863},[405],{"categories":1865},[110],{"categories":1867},[47],{"categories":1869},[47],{"categories":1871},[47],{"categories":1873},[],{"categories":1875},[134],{"categories":1877},[87],{"categories":1879},[],{"categories":1881},[92],{"categories":1883},[405],{"categories":1885},[],{"categories":1887},[131],{"categories":1889},[131],{"categories":1891},[],{"categories":1893},[141],{"categories":1895},[131],{"categories":1897},[47],{"categories":1899},[],{"categories":1901},[110],{"categories":1903},[47],{"categories":1905},[131],{"categories":1907},[92],{"categories":1909},[110],{"categories":1911},[],{"categories":1913},[92],{"categories":1915},[131],{"categories":1917},[47],{"categories":1919},[],{"categories":1921},[47],{"categories":1923},[47],{"categories":1925},[405],{"categories":1927},[110],{"categories":1929},[134],{"categories":1931},[134],{"categories":1933},[],{"categories":1935},[],{"categories":1937},[],{"categories":1939},[92],{"categories":1941},[141],{"categories":1943},[141],{"categories":1945},[],{"categories":1947},[],{"categories":1949},[47],{"categories":1951},[],{"categories":1953},[92],{"categories":1955},[47],{"categories":1957},[],{"categories":1959},[47],{"categories":1961},[87],{"categories":1963},[47],{"categories":1965},[148],{"categories":1967},[92],{"categories":1969},[47],{"categories":1971},[141],{"categories":1973},[110],{"categories":1975},[92],{"categories":1977},[],{"categories":1979},[110],{"categories":1981},[92],{"categories":1983},[92],{"categories":1985},[],{"categories":1987},[87],{"categories":1989},[92],{"categories":1991},[],{"categories":1993},[47],{"categories":1995},[84],{"categories":1997},[110],{"categories":1999},[405],{"categories":2001},[92],{"categories":2003},[92],{"categories":2005},[84],{"categories":2007},[47],{"categories":2009},[],{"categories":2011},[],{"categories":2013},[131],{"categories":2015},[47,87],{"categories":2017},[],{"categories":2019},[84],{"categories":2021},[134],{"categories":2023},[47],{"categories":2025},[141],{"categories":2027},[47],{"categories":2029},[92],{"categories":2031},[47],{"categories":2033},[47],{"categories":2035},[110],{"categories":2037},[92],{"categories":2039},[],{"categories":2041},[],{"categories":2043},[92],{"categories":2045},[47],{"categories":2047},[405],{"categories":2049},[],{"categories":2051},[47],{"categories":2053},[92],{"categories":2055},[],{"categories":2057},[47],{"categories":2059},[148],{"categories":2061},[134],{"categories":2063},[92],{"categories":2065},[47],{"categories":2067},[405],{"categories":2069},[],{"categories":2071},[47],{"categories":2073},[148],{"categories":2075},[131],{"categories":2077},[47],{"categories":2079},[],{"categories":2081},[148],{"categories":2083},[110],{"categories":2085},[47],{"categories":2087},[47],{"categories":2089},[84],{"categories":2091},[],{"categories":2093},[],{"categories":2095},[131],{"categories":2097},[47],{"categories":2099},[134],{"categories":2101},[148],{"categories":2103},[148],{"categories":2105},[110],{"categories":2107},[],{"categories":2109},[],{"categories":2111},[47],{"categories":2113},[],{"categories":2115},[47,141],{"categories":2117},[110],{"categories":2119},[92],{"categories":2121},[141],{"categories":2123},[47],{"categories":2125},[84],{"categories":2127},[],{"categories":2129},[],{"categories":2131},[84],{"categories":2133},[148],{"categories":2135},[47],{"categories":2137},[],{"categories":2139},[131,47],{"categories":2141},[405],{"categories":2143},[84],{"categories":2145},[],{"categories":2147},[87],{"categories":2149},[87],{"categories":2151},[47],{"categories":2153},[141],{"categories":2155},[92],{"categories":2157},[110],{"categories":2159},[148],{"categories":2161},[131],{"categories":2163},[47],{"categories":2165},[47],{"categories":2167},[47],{"categories":2169},[84],{"categories":2171},[47],{"categories":2173},[92],{"categories":2175},[110],{"categories":2177},[],{"categories":2179},[],{"categories":2181},[134],{"categories":2183},[141],{"categories":2185},[47],{"categories":2187},[131],{"categories":2189},[134],{"categories":2191},[47],{"categories":2193},[47],{"categories":2195},[92],{"categories":2197},[92],{"categories":2199},[47,87],{"categories":2201},[],{"categories":2203},[131],{"categories":2205},[],{"categories":2207},[47],{"categories":2209},[110],{"categories":2211},[84],{"categories":2213},[84],{"categories":2215},[92],{"categories":2217},[47],{"categories":2219},[87],{"categories":2221},[141],{"categories":2223},[148],{"categories":2225},[],{"categories":2227},[110],{"categories":2229},[47],{"categories":2231},[47],{"categories":2233},[110],{"categories":2235},[141],{"categories":2237},[47],{"categories":2239},[92],{"categories":2241},[110],{"categories":2243},[47],{"categories":2245},[131],{"categories":2247},[47],{"categories":2249},[47],{"categories":2251},[405],{"categories":2253},[95],{"categories":2255},[92],{"categories":2257},[47],{"categories":2259},[110],{"categories":2261},[92],{"categories":2263},[148],{"categories":2265},[47],{"categories":2267},[],{"categories":2269},[47],{"categories":2271},[],{"categories":2273},[],{"categories":2275},[],{"categories":2277},[87],{"categories":2279},[47],{"categories":2281},[92],{"categories":2283},[110],{"categories":2285},[110],{"categories":2287},[110],{"categories":2289},[110],{"categories":2291},[],{"categories":2293},[84],{"categories":2295},[92],{"categories":2297},[110],{"categories":2299},[84],{"categories":2301},[92],{"categories":2303},[47],{"categories":2305},[47,92],{"categories":2307},[92],{"categories":2309},[405],{"categories":2311},[110],{"categories":2313},[110],{"categories":2315},[92],{"categories":2317},[47],{"categories":2319},[],{"categories":2321},[110],{"categories":2323},[148],{"categories":2325},[84],{"categories":2327},[47],{"categories":2329},[47],{"categories":2331},[],{"categories":2333},[141],{"categories":2335},[],{"categories":2337},[84],{"categories":2339},[92],{"categories":2341},[110],{"categories":2343},[47],{"categories":2345},[110],{"categories":2347},[84],{"categories":2349},[110],{"categories":2351},[110],{"categories":2353},[],{"categories":2355},[87],{"categories":2357},[92],{"categories":2359},[110],{"categories":2361},[110],{"categories":2363},[110],{"categories":2365},[110],{"categories":2367},[110],{"categories":2369},[110],{"categories":2371},[110],{"categories":2373},[110],{"categories":2375},[110],{"categories":2377},[110],{"categories":2379},[134],{"categories":2381},[84],{"categories":2383},[47],{"categories":2385},[47],{"categories":2387},[],{"categories":2389},[47,84],{"categories":2391},[],{"categories":2393},[92],{"categories":2395},[110],{"categories":2397},[92],{"categories":2399},[47],{"categories":2401},[47],{"categories":2403},[47],{"categories":2405},[47],{"categories":2407},[47],{"categories":2409},[92],{"categories":2411},[87],{"categories":2413},[131],{"categories":2415},[110],{"categories":2417},[47],{"categories":2419},[],{"categories":2421},[],{"categories":2423},[92],{"categories":2425},[131],{"categories":2427},[47],{"categories":2429},[],{"categories":2431},[],{"categories":2433},[148],{"categories":2435},[47],{"categories":2437},[],{"categories":2439},[],{"categories":2441},[84],{"categories":2443},[87],{"categories":2445},[47],{"categories":2447},[87],{"categories":2449},[131],{"categories":2451},[],{"categories":2453},[110],{"categories":2455},[],{"categories":2457},[131],{"categories":2459},[47],{"categories":2461},[148],{"categories":2463},[],{"categories":2465},[148],{"categories":2467},[],{"categories":2469},[],{"categories":2471},[92],{"categories":2473},[],{"categories":2475},[87],{"categories":2477},[84],{"categories":2479},[131],{"categories":2481},[141],{"categories":2483},[],{"categories":2485},[],{"categories":2487},[47],{"categories":2489},[84],{"categories":2491},[148],{"categories":2493},[],{"categories":2495},[92],{"categories":2497},[92],{"categories":2499},[110],{"categories":2501},[47],{"categories":2503},[92],{"categories":2505},[47],{"categories":2507},[92],{"categories":2509},[47],{"categories":2511},[95],{"categories":2513},[110],{"categories":2515},[],{"categories":2517},[148],{"categories":2519},[141],{"categories":2521},[92],{"categories":2523},[],{"categories":2525},[47],{"categories":2527},[92],{"categories":2529},[87],{"categories":2531},[84],{"categories":2533},[47],{"categories":2535},[131],{"categories":2537},[141],{"categories":2539},[141],{"categories":2541},[47],{"categories":2543},[134],{"categories":2545},[47],{"categories":2547},[92],{"categories":2549},[87],{"categories":2551},[92],{"categories":2553},[47],{"categories":2555},[47],{"categories":2557},[92],{"categories":2559},[110],{"categories":2561},[],{"categories":2563},[84],{"categories":2565},[47],{"categories":2567},[92],{"categories":2569},[47],{"categories":2571},[47],{"categories":2573},[],{"categories":2575},[131],{"categories":2577},[87],{"categories":2579},[110],{"categories":2581},[47],{"categories":2583},[47],{"categories":2585},[131],{"categories":2587},[148],{"categories":2589},[134],{"categories":2591},[47],{"categories":2593},[110],{"categories":2595},[47],{"categories":2597},[92],{"categories":2599},[405],{"categories":2601},[47],{"categories":2603},[92],{"categories":2605},[134],{"categories":2607},[],{"categories":2609},[92],{"categories":2611},[141],{"categories":2613},[131],{"categories":2615},[47],{"categories":2617},[84],{"categories":2619},[87],{"categories":2621},[141],{"categories":2623},[],{"categories":2625},[92],{"categories":2627},[47],{"categories":2629},[],{"categories":2631},[110],{"categories":2633},[],{"categories":2635},[110],{"categories":2637},[47],{"categories":2639},[92],{"categories":2641},[92],{"categories":2643},[92],{"categories":2645},[],{"categories":2647},[],{"categories":2649},[47],{"categories":2651},[47],{"categories":2653},[],{"categories":2655},[131],{"categories":2657},[92],{"categories":2659},[148],{"categories":2661},[84],{"categories":2663},[],{"categories":2665},[],{"categories":2667},[110],{"categories":2669},[141],{"categories":2671},[47],{"categories":2673},[47],{"categories":2675},[47],{"categories":2677},[141],{"categories":2679},[110],{"categories":2681},[131],{"categories":2683},[47],{"categories":2685},[47],{"categories":2687},[47],{"categories":2689},[110],{"categories":2691},[47],{"categories":2693},[110],{"categories":2695},[92],{"categories":2697},[92],{"categories":2699},[141],{"categories":2701},[92],{"categories":2703},[47],{"categories":2705},[141],{"categories":2707},[131],{"categories":2709},[],{"categories":2711},[92],{"categories":2713},[],{"categories":2715},[],{"categories":2717},[],{"categories":2719},[87],{"categories":2721},[47],{"categories":2723},[92],{"categories":2725},[84],{"categories":2727},[92],{"categories":2729},[148],{"categories":2731},[],{"categories":2733},[92],{"categories":2735},[],{"categories":2737},[84],{"categories":2739},[92],{"categories":2741},[],{"categories":2743},[92],{"categories":2745},[47],{"categories":2747},[110],{"categories":2749},[47],{"categories":2751},[92],{"categories":2753},[110],{"categories":2755},[92],{"categories":2757},[141],{"categories":2759},[131],{"categories":2761},[84],{"categories":2763},[],{"categories":2765},[92],{"categories":2767},[131],{"categories":2769},[405],{"categories":2771},[110],{"categories":2773},[47],{"categories":2775},[131],{"categories":2777},[84],{"categories":2779},[],{"categories":2781},[92],{"categories":2783},[92],{"categories":2785},[47],{"categories":2787},[],{"categories":2789},[92],{"categories":2791},[95],{"categories":2793},[110],{"categories":2795},[92],{"categories":2797},[87],{"categories":2799},[],{"categories":2801},[47],{"categories":2803},[95],{"categories":2805},[47],{"categories":2807},[92],{"categories":2809},[110],{"categories":2811},[84],{"categories":2813},[405],{"categories":2815},[47],{"categories":2817},[47],{"categories":2819},[47],{"categories":2821},[110],{"categories":2823},[87],{"categories":2825},[47],{"categories":2827},[131],{"categories":2829},[110],{"categories":2831},[405],{"categories":2833},[47],{"categories":2835},[],{"categories":2837},[],{"categories":2839},[405],{"categories":2841},[134],{"categories":2843},[92],{"categories":2845},[92],{"categories":2847},[110],{"categories":2849},[47],{"categories":2851},[84],{"categories":2853},[131],{"categories":2855},[92],{"categories":2857},[47],{"categories":2859},[148],{"categories":2861},[47],{"categories":2863},[92],{"categories":2865},[],{"categories":2867},[47],{"categories":2869},[47],{"categories":2871},[110],{"categories":2873},[84],{"categories":2875},[],{"categories":2877},[47],{"categories":2879},[47],{"categories":2881},[141],{"categories":2883},[131],{"categories":2885},[47,92],{"categories":2887},[148,87],{"categories":2889},[47],{"categories":2891},[],{"categories":2893},[92],{"categories":2895},[],{"categories":2897},[141],{"categories":2899},[47],{"categories":2901},[110],{"categories":2903},[],{"categories":2905},[92],{"categories":2907},[],{"categories":2909},[131],{"categories":2911},[92],{"categories":2913},[84],{"categories":2915},[92],{"categories":2917},[47],{"categories":2919},[405],{"categories":2921},[148],{"categories":2923},[87],{"categories":2925},[87],{"categories":2927},[84],{"categories":2929},[84],{"categories":2931},[47],{"categories":2933},[92],{"categories":2935},[47],{"categories":2937},[47],{"categories":2939},[84],{"categories":2941},[47],{"categories":2943},[148],{"categories":2945},[110],{"categories":2947},[47],{"categories":2949},[92],{"categories":2951},[47],{"categories":2953},[],{"categories":2955},[141],{"categories":2957},[],{"categories":2959},[92],{"categories":2961},[84],{"categories":2963},[],{"categories":2965},[405],{"categories":2967},[47],{"categories":2969},[],{"categories":2971},[110],{"categories":2973},[92],{"categories":2975},[141],{"categories":2977},[47],{"categories":2979},[92],{"categories":2981},[141],{"categories":2983},[92],{"categories":2985},[110],{"categories":2987},[84],{"categories":2989},[110],{"categories":2991},[141],{"categories":2993},[47],{"categories":2995},[131],{"categories":2997},[47],{"categories":2999},[47],{"categories":3001},[47],{"categories":3003},[47],{"categories":3005},[92],{"categories":3007},[47],{"categories":3009},[92],{"categories":3011},[47],{"categories":3013},[84],{"categories":3015},[47],{"categories":3017},[92],{"categories":3019},[131],{"categories":3021},[84],{"categories":3023},[92],{"categories":3025},[131],{"categories":3027},[],{"categories":3029},[47],{"categories":3031},[47],{"categories":3033},[141],{"categories":3035},[],{"categories":3037},[92],{"categories":3039},[148],{"categories":3041},[47],{"categories":3043},[110],{"categories":3045},[148],{"categories":3047},[92],{"categories":3049},[87],{"categories":3051},[87],{"categories":3053},[47],{"categories":3055},[84],{"categories":3057},[],{"categories":3059},[47],{"categories":3061},[],{"categories":3063},[84],{"categories":3065},[47],{"categories":3067},[92],{"categories":3069},[92],{"categories":3071},[],{"categories":3073},[141],{"categories":3075},[141],{"categories":3077},[148],{"categories":3079},[131],{"categories":3081},[],{"categories":3083},[47],{"categories":3085},[84],{"categories":3087},[47],{"categories":3089},[141],{"categories":3091},[84],{"categories":3093},[110],{"categories":3095},[110],{"categories":3097},[],{"categories":3099},[110],{"categories":3101},[92],{"categories":3103},[131],{"categories":3105},[134],{"categories":3107},[47],{"categories":3109},[],{"categories":3111},[110],{"categories":3113},[141],{"categories":3115},[87],{"categories":3117},[47],{"categories":3119},[84],{"categories":3121},[405],{"categories":3123},[84],{"categories":3125},[],{"categories":3127},[],{"categories":3129},[110],{"categories":3131},[],{"categories":3133},[92],{"categories":3135},[92],{"categories":3137},[92],{"categories":3139},[],{"categories":3141},[47],{"categories":3143},[],{"categories":3145},[110],{"categories":3147},[84],{"categories":3149},[131],{"categories":3151},[47],{"categories":3153},[110],{"categories":3155},[110],{"categories":3157},[],{"categories":3159},[110],{"categories":3161},[84],{"categories":3163},[47],{"categories":3165},[],{"categories":3167},[92],{"categories":3169},[92],{"categories":3171},[84],{"categories":3173},[],{"categories":3175},[],{"categories":3177},[],{"categories":3179},[131],{"categories":3181},[92],{"categories":3183},[47],{"categories":3185},[],{"categories":3187},[],{"categories":3189},[],{"categories":3191},[131],{"categories":3193},[],{"categories":3195},[84],{"categories":3197},[],{"categories":3199},[],{"categories":3201},[131],{"categories":3203},[47],{"categories":3205},[110],{"categories":3207},[],{"categories":3209},[148],{"categories":3211},[110],{"categories":3213},[148],{"categories":3215},[47],{"categories":3217},[],{"categories":3219},[],{"categories":3221},[92],{"categories":3223},[],{"categories":3225},[],{"categories":3227},[92],{"categories":3229},[47],{"categories":3231},[],{"categories":3233},[92],{"categories":3235},[110],{"categories":3237},[148],{"categories":3239},[134],{"categories":3241},[92],{"categories":3243},[92],{"categories":3245},[],{"categories":3247},[],{"categories":3249},[],{"categories":3251},[110],{"categories":3253},[],{"categories":3255},[],{"categories":3257},[131],{"categories":3259},[84],{"categories":3261},[],{"categories":3263},[87],{"categories":3265},[148],{"categories":3267},[47],{"categories":3269},[141],{"categories":3271},[84],{"categories":3273},[134],{"categories":3275},[87],{"categories":3277},[141],{"categories":3279},[],{"categories":3281},[],{"categories":3283},[92],{"categories":3285},[84],{"categories":3287},[131],{"categories":3289},[84],{"categories":3291},[92],{"categories":3293},[405],{"categories":3295},[92],{"categories":3297},[],{"categories":3299},[47],{"categories":3301},[110],{"categories":3303},[141],{"categories":3305},[],{"categories":3307},[131],{"categories":3309},[110],{"categories":3311},[84],{"categories":3313},[92],{"categories":3315},[47],{"categories":3317},[87],{"categories":3319},[92,405],{"categories":3321},[92],{"categories":3323},[141],{"categories":3325},[47],{"categories":3327},[134],{"categories":3329},[148],{"categories":3331},[92],{"categories":3333},[],{"categories":3335},[92],{"categories":3337},[47],{"categories":3339},[87],{"categories":3341},[],{"categories":3343},[],{"categories":3345},[47],{"categories":3347},[134],{"categories":3349},[47],{"categories":3351},[],{"categories":3353},[110],{"categories":3355},[],{"categories":3357},[110],{"categories":3359},[141],{"categories":3361},[92],{"categories":3363},[47],{"categories":3365},[148],{"categories":3367},[141],{"categories":3369},[],{"categories":3371},[110],{"categories":3373},[47],{"categories":3375},[],{"categories":3377},[47],{"categories":3379},[92],{"categories":3381},[47],{"categories":3383},[92],{"categories":3385},[47],{"categories":3387},[47],{"categories":3389},[47],{"categories":3391},[47],{"categories":3393},[87],{"categories":3395},[],{"categories":3397},[95],{"categories":3399},[110],{"categories":3401},[47],{"categories":3403},[],{"categories":3405},[141],{"categories":3407},[47],{"categories":3409},[47],{"categories":3411},[92],{"categories":3413},[110],{"categories":3415},[47],{"categories":3417},[47],{"categories":3419},[87],{"categories":3421},[92],{"categories":3423},[131],{"categories":3425},[],{"categories":3427},[134],{"categories":3429},[47],{"categories":3431},[],{"categories":3433},[110],{"categories":3435},[148],{"categories":3437},[],{"categories":3439},[],{"categories":3441},[110],{"categories":3443},[110],{"categories":3445},[148],{"categories":3447},[84],{"categories":3449},[92],{"categories":3451},[92],{"categories":3453},[47],{"categories":3455},[87],{"categories":3457},[],{"categories":3459},[],{"categories":3461},[110],{"categories":3463},[134],{"categories":3465},[141],{"categories":3467},[92],{"categories":3469},[131],{"categories":3471},[134],{"categories":3473},[134],{"categories":3475},[],{"categories":3477},[110],{"categories":3479},[47],{"categories":3481},[47],{"categories":3483},[141],{"categories":3485},[],{"categories":3487},[110],{"categories":3489},[110],{"categories":3491},[110],{"categories":3493},[],{"categories":3495},[92],{"categories":3497},[47],{"categories":3499},[],{"categories":3501},[84],{"categories":3503},[87],{"categories":3505},[],{"categories":3507},[47],{"categories":3509},[47],{"categories":3511},[],{"categories":3513},[141],{"categories":3515},[],{"categories":3517},[],{"categories":3519},[],{"categories":3521},[],{"categories":3523},[47],{"categories":3525},[110],{"categories":3527},[],{"categories":3529},[],{"categories":3531},[47],{"categories":3533},[47],{"categories":3535},[47],{"categories":3537},[134],{"categories":3539},[47],{"categories":3541},[134],{"categories":3543},[],{"categories":3545},[134],{"categories":3547},[134],{"categories":3549},[405],{"categories":3551},[92],{"categories":3553},[141],{"categories":3555},[],{"categories":3557},[],{"categories":3559},[134],{"categories":3561},[141],{"categories":3563},[141],{"categories":3565},[141],{"categories":3567},[],{"categories":3569},[84],{"categories":3571},[141],{"categories":3573},[141],{"categories":3575},[84],{"categories":3577},[141],{"categories":3579},[87],{"categories":3581},[141],{"categories":3583},[141],{"categories":3585},[141],{"categories":3587},[134],{"categories":3589},[110],{"categories":3591},[110],{"categories":3593},[47],{"categories":3595},[141],{"categories":3597},[134],{"categories":3599},[405],{"categories":3601},[134],{"categories":3603},[134],{"categories":3605},[134],{"categories":3607},[],{"categories":3609},[87],{"categories":3611},[],{"categories":3613},[405],{"categories":3615},[141],{"categories":3617},[141],{"categories":3619},[141],{"categories":3621},[92],{"categories":3623},[110,87],{"categories":3625},[134],{"categories":3627},[],{"categories":3629},[],{"categories":3631},[134],{"categories":3633},[],{"categories":3635},[134],{"categories":3637},[110],{"categories":3639},[92],{"categories":3641},[],{"categories":3643},[141],{"categories":3645},[47],{"categories":3647},[131],{"categories":3649},[],{"categories":3651},[47],{"categories":3653},[],{"categories":3655},[110],{"categories":3657},[84],{"categories":3659},[134],{"categories":3661},[],{"categories":3663},[141],{"categories":3665},[110],[3667,3877,3976,4079],{"id":3668,"title":3669,"ai":3670,"body":3675,"categories":3860,"created_at":48,"date_modified":48,"description":3861,"extension":49,"faq":48,"featured":50,"kicker_label":48,"meta":3862,"navigation":63,"path":3863,"published_at":3864,"question":48,"scraped_at":3865,"seo":3866,"sitemap":3867,"source_id":3868,"source_name":3869,"source_type":3870,"source_url":3871,"stem":3872,"tags":3873,"thumbnail_url":48,"tldr":3874,"tweet":48,"unknown_tags":3875,"__hash__":3876},"summaries\u002Fsummaries\u002Ff932400d9db7252e-slash-llm-token-costs-10x-by-fixing-6-bad-habits-summary.md","Slash LLM Token Costs 10x by Fixing 6 Bad Habits",{"provider":7,"model":8,"input_tokens":3671,"output_tokens":3672,"processing_time_ms":3673,"cost_usd":3674},8213,2447,18362,0.00257525,{"type":14,"value":3676,"toc":3851},[3677,3681,3684,3691,3697,3701,3704,3709,3712,3717,3721,3724,3729,3732,3736,3739,3745,3748,3753,3757,3760,3766,3769,3774,3778,3781,3803,3806,3809,3814,3818],[17,3678,3680],{"id":3679},"file-formats-are-your-biggest-beginner-token-trap","File Formats Are Your Biggest Beginner Token Trap",[22,3682,3683],{},"Raw PDFs, images, and screenshots explode token counts because LLMs encode binary structure, headers, footers, fonts, and layout metadata. A newbie drags in three 1,500-word PDFs (4,500 words total) and asks Claude to \"Summarize these.\" What should be ~5,000 tokens balloons to 100,000+ due to formatting overhead. This waste compounds as the bloated context bounces back in every turn, filling your window fast.",[22,3685,3686,3690],{},[3687,3688,3689],"strong",{},"Fix:"," Convert to markdown first. Free web tools or a quick Claude prompt strips junk, yielding 4-6,000 clean tokens—a 20x saving. The speaker built a plugin for Open Brain ecosystem: ingest file, hit \"transform,\" get markdown. For 99% of cases, you only need text, not style. Trade-off: Lose visual fidelity, but gain speed and cost control. He calls file formats \"designed to be human readable, not AI readable.\"",[3692,3693,3694],"blockquote",{},[22,3695,3696],{},"\"4500 words of content can become a 100 plus thousand tokens if you're not careful all you have to do to avoid that is just think in terms of markdown... saving you 20x on the memory.\"\n(Context: Explaining PDF bloat; this quote shows why rookies hit limits in one chat.)",[17,3698,3700],{"id":3699},"conversation-sprawl-wastes-more-than-you-think","Conversation Sprawl Wastes More Than You Think",[22,3702,3703],{},"Intermediate users sprawl chats to 20-40 turns, diluting original instructions amid noise. Models compress history but still resend the full context each turn—every reply costs the entire prior exchange. Mixing research, ideation, and execution in one thread confuses the model and burns tokens.",[22,3705,3706,3708],{},[3687,3707,3689],{}," Separate modes. Use short, focused chats (10-15 turns max) for heavy work: gather intel in dedicated threads (Grok for X sentiment, ChatGPT for earnings, Perplexity for research, Claude for blogs), then synthesize in a final crisp prompt. Mark evolving chats upfront: \"Our goal is to evolve and conclude together.\" End with \"Summarize this.\" Start fresh often—long threads correlate with \"LLM psychosis\" as models drift.",[22,3710,3711],{},"Trade-off: More chats mean manual synthesis, but you avoid context dilution and get clearer outputs. Every turn resends history, so sprawling is like \"filling up the context window with croft.\"",[3692,3713,3714],{},[22,3715,3716],{},"\"Why make them suffer... why not just ask for what you want upfront... your objective... should be to be so clear that the AI needs to do nothing else and it just goes and gets the work done.\"\n(Context: Critiquing multi-mode sprawl; highlights single-turn design of LLMs.)",[17,3718,3720],{"id":3719},"plugins-and-preloads-the-silent-context-tax","Plugins and Preloads: The Silent Context Tax",[22,3722,3723],{},"Loading 10+ plugins (e.g., Google Drive you never use) adds 50,000+ tokens before you type—every chat. It's like dumping every workshop tool on the bench before picking a hammer. Hype drives additions, but they barnacle on forever.",[22,3725,3726,3728],{},[3687,3727,3689],{}," Audit ruthlessly. Use \u002Fcontext in Claude Code to check loads; disable unused connectors. Only equip 3-5 per task. For advanced setups, prune system prompts weekly—ditch lines from Claude 3.5 era.",[22,3730,3731],{},"Trade-off: Lose convenience for rarely used tools, but gain focus. Models pick wrong tools amid clutter.",[17,3733,3735],{"id":3734},"model-tiering-delivers-8-10x-savings-without-losing-quality","Model Tiering Delivers 8-10x Savings Without Losing Quality",[22,3737,3738],{},"Using Opus (or GPT-4o) for everything—formatting, proofreading, execution—is overkill. A production pipeline the speaker reviewed analyzes long conversations across dozens of dimensions on frontier models, yet costs \u003C25¢\u002Fuser because they tier: Opus for reasoning, Sonnet for execution, Haiku for polish.",[22,3740,3741,3744],{},[3687,3742,3743],{},"Example math (5-hour session, same output):"," Sloppy (raw PDFs, 30-turn sprawl, all-Opus): 800k-1M input tokens + 150-200k output = $8-10 ($5\u002FM input, $25\u002FM output). Clean (markdown, fresh chats every 10-15 turns, tiered models, scoped context): 100-150k input + 50-80k output = ~$1. Scale to 10-person team API: $2,000 vs. $250\u002Fmonth.",[22,3746,3747],{},"Trade-off: Test cheaper models per task; Haiku shines on polish but flops on complex reasoning. As models improve, lean out context—trust retrieval over frontloading.",[3692,3749,3750],{},[22,3751,3752],{},"\"Don't bring a Ferrari to the grocery store.\"\n(Context: Model tiering; punchy metaphor for using Opus everywhere.)",[17,3754,3756],{"id":3755},"production-levers-caching-search-and-auditing","Production Levers: Caching, Search, and Auditing",[22,3758,3759],{},"Advanced users screw up at scale (millions of tokens). Ignore prompt caching? Miss 90% discounts (Opus: $0.50\u002FM cached vs. $5\u002FM). System prompts bloat from unpruned cruft. Web search via native Claude burns 10-50k tokens\u002Fquery vs. Perplexity (5x faster, structured citations).",[22,3761,3762,3765],{},[3687,3763,3764],{},"Fixes:"," Cache stable context (prompts, tools, docs). Use MCP connectors for cheap search (e.g., Perplexity service). For agents\u002Frepos, test context needs per model gen—dumber models needed fat windows; now trim.",[22,3767,3768],{},"Jen-Hsun Huang pegs engineer token spend at $250k\u002Fyear—don't be that person. With Mythos\u002FGPT-next\u002FGemini (GB300-trained, 10x Opus pricing rumored: $50\u002FM in, $250\u002FM out), sloppy habits scale painfully.",[3692,3770,3771],{},[22,3772,3773],{},"\"The models are not expensive it's your habits that cost a lot... your mistakes scale with the price of intelligence.\"\n(Context: Thesis opener and closer; frames costs as behavioral, not inherent.)",[17,3775,3777],{"id":3776},"tools-to-diagnose-and-fix-your-usage","Tools to Diagnose and Fix Your Usage",[22,3779,3780],{},"Speaker built a \"stupid button\" (Open Brain plugin\u002Fskill\u002Fguardrails):",[3782,3783,3784,3791,3797],"ol",{},[3785,3786,3787,3790],"li",{},[3687,3788,3789],{},"Audit prompt:"," Paste recent chat; flags raw docs, sprawl, model misuse, redundant loads—prioritizes fixes.",[3785,3792,3793,3796],{},[3687,3794,3795],{},"Gas tank skill:"," Measures per-session overhead (system prompts, plugins); before\u002Fafter baselines.",[3785,3798,3799,3802],{},[3687,3800,3801],{},"Guardrails:"," Blocks token-waste on knowledge stores.",[22,3804,3805],{},"Run it: Answers 6 questions (raw files? Fresh chats? All-Opus? Preloads? Caching? Cheap search?). No setup for prompt version.",[22,3807,3808],{},"Real pipeline proves frontier AI viability: Dozens of analysis dimensions on long convos, personalized output, \u003C25¢\u002Fuser.",[3692,3810,3811],{},[22,3812,3813],{},"\"Frontier AI can be absurdly cheap when you know what you're doing... most of us are spending more than we need to on AI.\"\n(Context: Production example; counters \"AI is too expensive\" narrative.)",[17,3815,3817],{"id":3816},"key-takeaways","Key Takeaways",[3819,3820,3821,3824,3827,3830,3833,3836,3839,3842,3845,3848],"ul",{},[3785,3822,3823],{},"Convert all inputs to markdown: 20x token savings on docs\u002Fimages; use free tools or Claude.",[3785,3825,3826],{},"Cap chats at 10-15 turns; separate research from execution for clarity and cost.",[3785,3828,3829],{},"Audit plugins\u002Fpreloads weekly: Disable barnacles adding 50k+ tokens\u002Fchat.",[3785,3831,3832],{},"Tier models: Opus reasoning, Sonnet execution, Haiku polish—8-10x cheaper same output.",[3785,3834,3835],{},"Cache stable context: 90% off repeated inputs; essential for agents\u002Fproduction.",[3785,3837,3838],{},"Use cheap search (Perplexity\u002FMCP): 10-50k fewer tokens\u002Fquery, faster results.",[3785,3840,3841],{},"Prune system prompts biweekly; trim context as models smarten.",[3785,3843,3844],{},"Baseline usage with audits: Turn $10\u002Fday slop into $1\u002Fday efficiency.",[3785,3846,3847],{},"Prep for 10x pricier models: Habits today dictate ROI tomorrow.",[3785,3849,3850],{},"Build token smarts: $250k\u002Fyear engineer spend is avoidable skill gap.",{"title":40,"searchDepth":41,"depth":41,"links":3852},[3853,3854,3855,3856,3857,3858,3859],{"id":3679,"depth":41,"text":3680},{"id":3699,"depth":41,"text":3700},{"id":3719,"depth":41,"text":3720},{"id":3734,"depth":41,"text":3735},{"id":3755,"depth":41,"text":3756},{"id":3776,"depth":41,"text":3777},{"id":3816,"depth":41,"text":3817},[47],"My site: https:\u002F\u002Fnatebjones.com\nFull Story w\u002F Prompts: https:\u002F\u002Fnatesnewsletter.substack.com\u002Fp\u002Fyour-claude-sessions-cost-10x-what?r=1z4sm5&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true\n___________________\nWhat's really happening inside your AI costs when Jensen Hwang says engineers will spend $250,000 a year on tokens?\n\nThe common story is that frontier models are expensive — but the reality is that your habits cost more than the models ever will, and most users burn 8-10x what they need to.\n\nIn this video, I share the inside scoop on token efficiency before Mythos pricing hits:\n\n • Why raw PDFs can turn 4,500 words into 100,000 tokens\n • How conversation sprawl compounds waste with every turn\n • What plugin overhead costs you before you type a word\n • Where model mixing drops a $10 session to $1\n\nBuilders who keep burning tokens as a badge of honor will face a reckoning when cutting-edge models cost 10x what Opus costs today — the habits you build now determine whether you scale or stall.\n\nChapters\n00:00 Stop burning tokens and blaming the model\n02:30 A real pipeline that costs less than 25 cents per user\n04:30 Rookie mistake: document ingestion and PDFs\n07:00 Convert to Markdown, always\n09:00 Conversation sprawl and context compression\n11:30 The plugin and connector tax\n14:00 Advanced users have the most expensive mistakes\n16:30 The 8-10x cost reduction breakdown\n19:00 What Mythos pricing will do to your mistakes\n21:00 The stupid button: six questions to audit yourself\n23:30 Five commandments for agent token management\n26:00 Use your tokens well, not wastefully\n\nSubscribe for daily AI strategy and news.\nFor deeper playbooks and analysis: https:\u002F\u002Fnatesnewsletter.substack.com\u002F\n\nListen to this video as a podcast.\n- Spotify: https:\u002F\u002Fopen.spotify.com\u002Fshow\u002F0gkFdjd1wptEKJKLu9LbZ4\n- Apple Podcasts: https:\u002F\u002Fpodcasts.apple.com\u002Fus\u002Fpodcast\u002Fai-news-strategy-daily-with-nate-b-jones\u002Fid1877109372",{},"\u002Fsummaries\u002Ff932400d9db7252e-slash-llm-token-costs-10x-by-fixing-6-bad-habits-summary","2026-04-02 14:00:06","2026-04-03 21:11:38",{"title":3669,"description":3861},{"loc":3863},"f932400d9db7252e","AI News & Strategy Daily | Nate B Jones","video","https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=5ztI_dbj6ek","summaries\u002Ff932400d9db7252e-slash-llm-token-costs-10x-by-fixing-6-bad-habits-summary",[75,76,77],"Upcoming frontier models like Claude Mythos will cost 10x more—fix habits like raw PDFs, conversation sprawl, and overusing Opus to drop daily costs from $10 to $1 while getting the same output.",[77],"P_zaT8VyQum6hcxq8OX28jk4OI1NmCnKGvRnX42P8CU",{"id":3878,"title":3879,"ai":3880,"body":3885,"categories":3946,"created_at":48,"date_modified":48,"description":40,"extension":49,"faq":48,"featured":50,"kicker_label":48,"meta":3947,"navigation":63,"path":3962,"published_at":3963,"question":48,"scraped_at":3964,"seo":3965,"sitemap":3966,"source_id":3967,"source_name":3968,"source_type":71,"source_url":3969,"stem":3970,"tags":3971,"thumbnail_url":48,"tldr":3973,"tweet":48,"unknown_tags":3974,"__hash__":3975},"summaries\u002Fsummaries\u002Fa99451de2de64e60-claude-md-patterns-for-bulletproof-ai-coding-summary.md","Claude.md Patterns for Bulletproof AI Coding",{"provider":7,"model":8,"input_tokens":3881,"output_tokens":3882,"processing_time_ms":3883,"cost_usd":3884},7402,1645,34494,0.00179605,{"type":14,"value":3886,"toc":3941},[3887,3891,3894,3897,3900,3903,3906,3910,3913,3916,3919,3922,3925,3929,3932,3935,3938],[17,3888,3890],{"id":3889},"karpathy-inspired-rules-to-align-claude-with-your-intent","Karpathy-Inspired Rules to Align Claude with Your Intent",[22,3892,3893],{},"Start every claude.md with a project description at the top so Claude grasps the app's structure, services, dependencies, and runtime before diving in—this prevents deduction from code alone and cuts misalignment. Add explicit 'think before coding': Claude must state assumptions, list multiple interpretations if ambiguous, and confirm your choice, slashing course-corrections by forcing clarification over training-data guesses.",[22,3895,3896],{},"Prioritize simplicity: Instruct Claude to solve in minimal lines (e.g., refactor if >200 lines when 50 suffice), add only requested features with proper error handling, and iterate toward conciseness. This avoids verbose overhead that bloats tokens, delays refactoring, and hinders scaling in large apps.",[22,3898,3899],{},"Enforce surgical changes: Touch only task-tracing code; flag unrelated issues (dead code, formatting) without fixing unless asked, as agents scatter focus on 'improvements.' Every edit must link directly to your request, listing other findings for your triage.",[22,3901,3902],{},"Drive goal execution: For each task, Claude defines verifiable success criteria upfront—like writing passing tests for validation inputs\u002Foutputs—then plans, implements, iterates until verified. For UI, pair with tools like Claude Chrome extension or Puppeteer MCP to visually confirm changes, as code alone misleads.",[22,3904,3905],{},"These patterns from Andrej Karpathy's skills repo transform vague tasks into precise, testable outcomes, ensuring behavior matches intent without wild implementations.",[17,3907,3909],{"id":3908},"tool-overrides-safety-and-iterative-refinement","Tool Overrides, Safety, and Iterative Refinement",[22,3911,3912],{},"Override defaults: Skip init-generated commands (e.g., npm run dev) Claude already knows; specify custom tools like GitHub CLI over git, PNPM over npm, or non-standard run instructions to leverage your stack without fallbacks.",[22,3914,3915],{},"Update dynamically: After user corrections, Claude applies fixes then logs learnings to a dedicated file, building a knowledge base of pitfalls and preferences for future tasks—treat claude.md as living, not static.",[22,3917,3918],{},"Embed git safety: Ban irreversible commands (force-push, reset --hard, rm -rf) without confirmation; if unsure, always ask. This guards production from accidents like unwanted merges.",[22,3920,3921],{},"Use path-scoped rule files: Create e.g., api-rules.md (first line declares scope) for file-type rules, referenced in root claude.md. This avoids bloat—Claude loads only relevant rules, staying focused without interference.",[22,3923,3924],{},"For monorepos, add scoped claude.md per subfolder for module-specific guidance; root holds global rules only, preventing divergence from irrelevant instructions.",[17,3926,3928],{"id":3927},"prioritized-structure-and-verification-for-peak-performance","Prioritized Structure and Verification for Peak Performance",[22,3930,3931],{},"Order by priority: Hard rules first (non-negotiable, e.g., safety, scoping), then medium (key principles like simplicity), finally low (references). Burying criticals dilutes impact.",[22,3933,3934],{},"Mandate full verification before completion: Don't just add code—run builds, tests, linting, type checks to confirm functionality. Report only when all pass, using every mechanism for fidelity.",[22,3936,3937],{},"Cap at 300 lines: Beyond this, performance drops; trim ruthlessly for focus.",[22,3939,3940],{},"This setup, refined from community testing and shipping, eliminates agent fights: Claude reasons correctly, changes precisely, verifies rigorously, and adapts—saving hours on real projects.",{"title":40,"searchDepth":41,"depth":41,"links":3942},[3943,3944,3945],{"id":3889,"depth":41,"text":3890},{"id":3908,"depth":41,"text":3909},{"id":3927,"depth":41,"text":3928},[47],{"content_references":3948,"triage":3960},[3949,3955],{"type":3950,"title":3951,"author":3952,"url":3953,"context":3954},"other","andrej-karpathy-skills","forrestchang","https:\u002F\u002Fgithub.com\u002Fforrestchang\u002Fandrej-karpathy-skills\u002F","cited",{"type":3956,"title":3957,"url":3958,"context":3959},"tool","Klaus","https:\u002F\u002Fklausai.com\u002Fr\u002FMv1e2","mentioned",{"relevance":59,"novelty":60,"quality":60,"actionability":59,"composite":61,"reasoning":3961},"Category: AI & LLMs. The article provides practical patterns for using Claude.md effectively, addressing the pain points of AI-Curious Developers and Technical Founders by offering concrete strategies for coding with AI. It emphasizes actionable steps like starting with a project description and defining success criteria, making it immediately applicable for building AI-powered products.","\u002Fsummaries\u002Fa99451de2de64e60-claude-md-patterns-for-bulletproof-ai-coding-summary","2026-04-28 14:30:29","2026-05-03 16:44:39",{"title":3879,"description":40},{"loc":3962},"c6527f0f4e352415","AI LABS","https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=fMY5Sdj2DMk","summaries\u002Fa99451de2de64e60-claude-md-patterns-for-bulletproof-ai-coding-summary",[75,3972,76,77],"agents","Craft claude.md with project description first, Karpathy rules like 'think before coding' and simplicity, tool overrides, git safety, scoped files, verification steps, and priority-ordered instructions under 300 lines to make Claude ship exact implementations without guesswork or bloat.",[77],"aSUZZ1N9_uYmZbJKhVlN4jk82uo4tI6cro4gCyHSS4w",{"id":3977,"title":3978,"ai":3979,"body":3984,"categories":4042,"created_at":48,"date_modified":48,"description":40,"extension":49,"faq":48,"featured":50,"kicker_label":48,"meta":4043,"navigation":63,"path":4066,"published_at":4067,"question":48,"scraped_at":4068,"seo":4069,"sitemap":4070,"source_id":4071,"source_name":4072,"source_type":71,"source_url":4073,"stem":4074,"tags":4075,"thumbnail_url":48,"tldr":4076,"tweet":48,"unknown_tags":4077,"__hash__":4078},"summaries\u002Fsummaries\u002F9c9d012e4625ef48-fix-claude-code-for-opus-4-7-9-key-changes-summary.md","Fix Claude Code for Opus 4.7: 9 Key Changes",{"provider":7,"model":8,"input_tokens":3980,"output_tokens":3981,"processing_time_ms":3982,"cost_usd":3983},6575,2115,10779,0.00186565,{"type":14,"value":3985,"toc":4036},[3986,3990,3993,3996,4000,4014,4018,4026,4030,4033],[17,3987,3989],{"id":3988},"adopt-ex-high-effort-and-adaptive-thinking-as-defaults","Adopt Ex-High Effort and Adaptive Thinking as Defaults",[22,3991,3992],{},"Opus 4.7 introduces ex-high effort level between high (4.6 default) and max, recommended verbatim by Anthropic as the new starting point for coding and agentic tasks since max yields diminishing returns and overthinking. Set max output tokens to 64,000 at ex-high or max to give the model room to think and act. Raise effort instead of rewriting shallow prompts. Adaptive thinking replaces extended thinking (which now returns HTTP 400 error on fixed budget_tokens); Claude auto-decides reasoning depth, outperforming the old mode per Anthropic evals—steer with pros like \"think carefully step by step\" for hard problems or \"prioritize quick response\" for easy ones. Tokenizer change uses ~1.35x more tokens for same text, filling context faster. Sampling params (temperature, top_p\u002Fk) also 400-error.",[22,3994,3995],{},"Treat Claude as capable engineer, not pair programmer: delegate via first-message structure stating intent, desired outcome, constraints, acceptance criteria, and file locations to minimize per-turn reasoning costs from ambiguity. Test prompts on colleagues—if confusing to them, confusing to Claude.",[17,3997,3999],{"id":3998},"build-5-layer-prompts-and-explicit-verbs-for-30-quality-gains","Build 5-Layer Prompts and Explicit Verbs for 30% Quality Gains",[22,4001,4002,4003],{},"Use one precise verb per instruction since 4.7 follows literally: \"suggest changes\" yields list only, no code; \"change this function\" edits directly. Layer prompts as: 1) clear\u002Fdirect instructions, 2) context\u002Fexplanation of why, 3) 3-5 XML-tagged examples (multi-shot beats single-shot), 4) XML structure (",[4004,4005,4006,4007],"instructions",{},", ",[4008,4009,4006,4010,4013],"context",{},[4011,4012],"input",{},"), 5) system role. Place long documents at prompt top, question at bottom for +30% response quality on multi-document inputs. For subagents, explicitly request \"spawn multiple sub-agents in the same turn\" when fanning out across files\u002Fbranches; skip for single-file work to avoid trash.",[17,4015,4017],{"id":4016},"enforce-context-hygiene-claudemd-and-hooks-for-reliability","Enforce Context Hygiene, CLAUDE.md, and Hooks for Reliability",[22,4019,4020,4021,4025],{},"Context window degrades performance: 4.5-minute fresh tasks stretch to 18 minutes after autos. Use \u002Fclear post-resolution, \u002Frewind on wrong turns. CLAUDE.md (auto-init via ",[4022,4023,4024],"code",{},"claude \u002Finit",") reads first each session with hierarchy: org policy > ~\u002F.claude user prefs > project-level > local gitignored > path-scoped rules—keep concise, cut ignorable lines to avoid bloat. Hooks via settings.json PreToolUse matchers block destructives pre-execution: command (shell script), HTTP (team endpoint), prompt (LM eval), agent (sub-agent verify).",[17,4027,4029],{"id":4028},"use-filesystem-memory-and-verification-for-long-tasks","Use Filesystem Memory and Verification for Long Tasks",[22,4031,4032],{},"Highest-leverage fix (per docs): always give Claude self-verification like tests\u002Fscreenshots\u002Fexpected output, never speculate on unopened code. For long-horizon: filesystem as memory via tests.json (passing\u002Ffailing\u002Fpending, never edit\u002Fremove), progress.txt notes, git commits as checkpoints. Code reviews: drop \"high-severity only\" filters (suppresses 11% bug-finding gains); use \"report every issue, even uncertain—filter later.\"",[22,4034,4035],{},"Checklist: 1) First-turn intent\u002Fconstraints\u002Facceptance, 2) ex-high default, 3) adaptive thinking, 4) CLAUDE.md, 5) pre-tool hook.",{"title":40,"searchDepth":41,"depth":41,"links":4037},[4038,4039,4040,4041],{"id":3988,"depth":41,"text":3989},{"id":3998,"depth":41,"text":3999},{"id":4016,"depth":41,"text":4017},{"id":4028,"depth":41,"text":4029},[47],{"content_references":4044,"triage":4064},[4045,4049,4052,4055,4058,4061],{"type":3950,"title":4046,"author":4047,"url":4048,"context":3954},"Best practices for using Claude Opus 4.7 with Claude Code","Anthropic","https:\u002F\u002Fclaude.com\u002Fblog\u002Fbest-practices-for-using-claude-opus-4-7-with-claude-code",{"type":3950,"title":4050,"url":4051,"context":3954},"Claude Code docs","https:\u002F\u002Fdocs.claude.com\u002Fen\u002Fdocs\u002Fclaude-code",{"type":3950,"title":4053,"url":4054,"context":3954},"Claude Code best-practices engineering page","https:\u002F\u002Fwww.anthropic.com\u002Fengineering\u002Fclaude-code-best-practices",{"type":3950,"title":4056,"url":4057,"context":3959},"Anthropic docs home","https:\u002F\u002Fdocs.anthropic.com",{"type":3950,"title":4059,"url":4060,"context":3959},"Boris Cherny on X","https:\u002F\u002Fx.com\u002Fbcherny",{"type":3956,"title":4062,"url":4063,"context":3959},"diy-yt-creator","https:\u002F\u002Fgithub.com\u002FLeex279\u002Fdiy-yt-creator",{"relevance":59,"novelty":60,"quality":60,"actionability":59,"composite":61,"reasoning":4065},"Category: AI & LLMs. The article provides specific, actionable strategies for optimizing the use of the Claude model in coding tasks, addressing pain points related to prompt engineering and production readiness. It details new defaults and prompt structures that can directly enhance productivity, making it highly relevant for developers integrating AI into their workflows.","\u002Fsummaries\u002F9c9d012e4625ef48-fix-claude-code-for-opus-4-7-9-key-changes-summary","2026-04-20 12:18:23","2026-04-21 15:22:05",{"title":3978,"description":40},{"loc":4066},"7affeb01a39ba436","DIY Smart Code","https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=0x5-XD9XD6c","summaries\u002F9c9d012e4625ef48-fix-claude-code-for-opus-4-7-9-key-changes-summary",[75,76,3972,77],"Opus 4.7 boosts coding power 13% but breaks old prompts—default to ex-high effort, adaptive thinking, literal verbs, and verification to resolve 3x more production tasks.",[77],"eHv33wsVFa7bYsq2egrHJ21E2-RW5c6qab7gKdTQUuc",{"id":4080,"title":4081,"ai":4082,"body":4087,"categories":4197,"created_at":48,"date_modified":48,"description":40,"extension":49,"faq":48,"featured":50,"kicker_label":48,"meta":4198,"navigation":63,"path":4208,"published_at":4209,"question":48,"scraped_at":4210,"seo":4211,"sitemap":4212,"source_id":4213,"source_name":4214,"source_type":71,"source_url":4215,"stem":4216,"tags":4217,"thumbnail_url":48,"tldr":4218,"tweet":48,"unknown_tags":4219,"__hash__":4220},"summaries\u002Fsummaries\u002F172e69741ae7f77d-vs-code-agent-loop-tools-sub-agents-and-optimizati-summary.md","VS Code Agent Loop: Tools, Sub-Agents, and Optimizations",{"provider":7,"model":8,"input_tokens":4083,"output_tokens":4084,"processing_time_ms":4085,"cost_usd":4086},8899,2050,10763,0.00251365,{"type":14,"value":4088,"toc":4190},[4089,4093,4096,4104,4107,4111,4114,4121,4124,4128,4131,4134,4137,4140,4144,4147,4150,4153,4156,4158],[17,4090,4092],{"id":4091},"agent-loop-fundamentals","Agent Loop Fundamentals",[22,4094,4095],{},"Pierce Boggan explains the agent loop as a giant while loop triggered by a user's first prompt in VS Code's GitHub Copilot chat. Each iteration sends an API request to a model with dynamically built components: a system prompt tailored to the selected model family (optimized pre- and post-launch via A\u002FB tests and evaluations), explicit context (e.g., mentioned files like hello.tsx), implicit context (open editors, running terminals, environment info), available tools, and the user prompt.",[22,4097,4098,4099,4103],{},"Tools form the core: unlike basic chat's text-only responses, agents choose from built-in tools (search, read\u002Fedit files, GitHub NCP servers) or custom ones, each with schemas and descriptions. The model decides actions—like searching files, reading them, then editing—appending outputs to iterate until issuing a stop message with a user summary. \"Imagine you just basically have a giant while loop... every ",[4100,4101,4102],"span",{},"interaction is"," an API request to a model.\"",[22,4105,4106],{},"James Monte Magno notes the loop's evolution over 6-8 months, with growing options like bypass, autopilot, planning modes, custom agents, and reasoning levels. Users see branches in chat (e.g., research via grepping\u002Freading files), all driven by the model appending prior outputs as context.",[17,4108,4110],{"id":4109},"harness-optimizations-and-model-tuning","Harness Optimizations and Model Tuning",[22,4112,4113],{},"The \"harness\"—prompts, context gathering, tools, and custom backend models—differentiates VS Code from CLI or other agents. Pierce highlights massive unseen optimization: VS Code's team (15-20 people) refines prompts with providers like Anthropic, OpenAI, Gemini, xAI weeks\u002Fmonths pre-launch using VS SWE-bench (custom, pollution-free alternative to SWE-bench). They analyze agent trajectories—not just pass\u002Ffail, but optimal paths for faster resolutions (1 minute vs. 1 hour).",[22,4115,4116,4117,4120],{},"Post-launch, A\u002FB tests and online evals handle demand spikes (e.g., new Opus 4.7 launch day capacity issues). Results: from 52-53% GPT-4o code commit rate to 90% with o1. Custom models tackle specifics, like agentic code retrieval for edits or cheap models for chat titles. \"With Opus ",[4100,4118,4119],{},"o1",", we're getting 90% of Opus code in our harness committed... improvement we see in 1 year.\"",[22,4122,4123],{},"Pierce stresses continuous loops: model updates, prompt\u002Ftool tweaks, purpose-built models. New models start \"infant\" but hone quickly; different models (4.5 to 4.7, o1 to 5.3 Codex) think differently, requiring per-model tuning.",[17,4125,4127],{"id":4126},"sub-agents-specialization-and-model-choices","Sub-Agents: Specialization and Model Choices",[22,4129,4130],{},"Sub-agents address when the main agent delegates: it's a tool the model selects to run a fresh agent loop with a goal, returning results like a function. Users question different models (e.g., main o1-preview at 3x cost, sub-agent Haiku at 0.33x): no bait-and-switch, but deliberate for best experience.",[22,4132,4133],{},"Reasons: speed\u002Fcost for narrow tasks (context gathering, exploration). Main agent (heavy reasoning model) plans\u002Fcoordinates; sub-agents use fast\u002Fcheaper models. Pierce: sub-agent as \"run this workflow with fresh context... return back to main thread.\" Model decides via tool choice in the loop.",[22,4135,4136],{},"Customizations modify basics: instructions append text (global or glob-patterned), skills let model fetch\u002Fappend context like tools, NCP adds tools. Trade-offs abound: too many tools\u002Foptions degrade choice (like humans with overload); custom models prune to relevant ones. User corrections append as text, enabling smart pivots but risking bad paths—kill\u002Frestart advised.",[22,4138,4139],{},"\"When you give people more choices, their ability to pick the right choice degrades.\"",[17,4141,4143],{"id":4142},"trade-offs-in-customization-and-behaviors","Trade-Offs in Customization and Behaviors",[22,4145,4146],{},"Pierce warns against extremes: stuffing prompts fills context windows; 1,000 tools overwhelm. Optimizations include tool-refining models and context-specific custom models. Bad loops from poor prior tokens require intervention, as each predicts the next.",[22,4148,4149],{},"Features like auto-titles, commit messages, PRs, next edits run mini-loops transparently. Harness tailors to code quality; incentives align to user success, not tricks. James emphasizes micro-decisions' impact on prompting.",[22,4151,4152],{},"Ongoing: demand prediction challenges in agentic era (10+ parallel agents), offline eval limits, provider updates.",[22,4154,4155],{},"\"There's an enormous amount of optimization... that you don't actually see.\"",[17,4157,3817],{"id":3816},[3819,4159,4160,4163,4166,4169,4172,4175,4178,4181,4184,4187],{},[3785,4161,4162],{},"Trigger the agent loop with a clear prompt; watch iterations via chat for search\u002Fread\u002Fedit patterns.",[3785,4164,4165],{},"Select models wisely—new ones like o1-preview need weeks to optimize; expect initial capacity hiccups.",[3785,4167,4168],{},"Use instructions\u002Fskills sparingly to avoid context bloat; let model choose via tools.",[3785,4170,4171],{},"Kill bad sub-agent paths early—corrections append as text, but prior tokens influence heavily.",[3785,4173,4174],{},"Customize via glob instructions or NCP for targeted tools, but limit options to aid model decisions.",[3785,4176,4177],{},"Evaluate via trajectories, not just resolution: aim for optimal paths in your workflows.",[3785,4179,4180],{},"Leverage VS SWE-bench insights: focus on production harnesses over polluted benchmarks like SWE-bench.",[3785,4182,4183],{},"For sub-agents, embrace model mixing—cheap\u002Ffast for exploration, heavy for orchestration.",[3785,4185,4186],{},"Monitor trade-offs: more tools degrade choice; use custom models for retrieval\u002Fedits.",[3785,4188,4189],{},"Stay updated weekly—harness evolves with models, boosting code acceptance from ~50% to 90%.",{"title":40,"searchDepth":41,"depth":41,"links":4191},[4192,4193,4194,4195,4196],{"id":4091,"depth":41,"text":4092},{"id":4109,"depth":41,"text":4110},{"id":4126,"depth":41,"text":4127},{"id":4142,"depth":41,"text":4143},{"id":3816,"depth":41,"text":3817},[47],{"content_references":4199,"triage":4205},[4200,4202],{"type":3950,"title":4201,"context":3959},"SWE-bench",{"type":3950,"title":4203,"author":4204,"context":3959},"VS SWE-bench","VS Code team",{"relevance":59,"novelty":60,"quality":60,"actionability":60,"composite":4206,"reasoning":4207},4.35,"Category: AI & LLMs. The article provides in-depth insights into the workings of VS Code's agent loop, addressing practical applications of AI in software development, which is highly relevant for the target audience. It discusses optimizations and model tuning that can directly impact developer productivity, making it actionable for those looking to implement similar strategies.","\u002Fsummaries\u002F172e69741ae7f77d-vs-code-agent-loop-tools-sub-agents-and-optimizati-summary","2026-04-20 07:00:11","2026-04-21 15:18:05",{"title":4081,"description":40},{"loc":4208},"dc097aac623090d9","Visual Studio Code","https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=ENxVTtLW_Bc","summaries\u002F172e69741ae7f77d-vs-code-agent-loop-tools-sub-agents-and-optimizati-summary",[3972,76,75,77],"VS Code's agent loop is a dynamic while loop powered by model-tuned prompts, context gathering, and tools; sub-agents use cheaper models for speed, with constant harness optimizations boosting code quality from 53% to 90%.",[77],"uImXX1ennm8fGVPFPDPsEOiC9Gh5g6ByReGzYV77Ruc"]