| 1 | { |
| 2 | "about": "A handful of tasks shaped like g1t's own agent work, for estimating what routing saves. Illustrative, not production data: tokens follow the profile in docs/research/reviewbench.md (turns from the change's size, 92% of input read from cache, output per turn), and failsOn is an assumption, the tiers a task of that difficulty is expected to fail on. Run with --live for g1t's real runs.", |
| 3 | "tasks": [ |
| 4 | { |
| 5 | "id": "fix-typo-in-readme", |
| 6 | "kind": "implement", |
| 7 | "change": { |
| 8 | "files": 1, |
| 9 | "lines": 6, |
| 10 | "sensitive": [] |
| 11 | }, |
| 12 | "labels": [ |
| 13 | "docs" |
| 14 | ], |
| 15 | "merged": true, |
| 16 | "tokens": { |
| 17 | "input": 11455, |
| 18 | "output": 8400, |
| 19 | "cacheRead": 351314, |
| 20 | "cacheWrite": 48072 |
| 21 | } |
| 22 | }, |
| 23 | { |
| 24 | "id": "add-search-filter", |
| 25 | "kind": "implement", |
| 26 | "change": { |
| 27 | "files": 6, |
| 28 | "lines": 240, |
| 29 | "sensitive": [] |
| 30 | }, |
| 31 | "merged": true, |
| 32 | "tokens": { |
| 33 | "input": 80004, |
| 34 | "output": 31200, |
| 35 | "cacheRead": 2453474, |
| 36 | "cacheWrite": 118380 |
| 37 | } |
| 38 | }, |
| 39 | { |
| 40 | "id": "rework-billing-ledger", |
| 41 | "kind": "implement", |
| 42 | "change": { |
| 43 | "files": 14, |
| 44 | "lines": 900, |
| 45 | "sensitive": [] |
| 46 | }, |
| 47 | "failsOn": [ |
| 48 | "small", |
| 49 | "large" |
| 50 | ], |
| 51 | "merged": true, |
| 52 | "tokens": { |
| 53 | "input": 184590, |
| 54 | "output": 48000, |
| 55 | "cacheRead": 5660760, |
| 56 | "cacheWrite": 178800 |
| 57 | } |
| 58 | }, |
| 59 | { |
| 60 | "id": "split-auth-service", |
| 61 | "kind": "implement", |
| 62 | "change": { |
| 63 | "files": 20, |
| 64 | "lines": 1500, |
| 65 | "sensitive": [] |
| 66 | }, |
| 67 | "labels": [ |
| 68 | "architecture" |
| 69 | ], |
| 70 | "failsOn": [ |
| 71 | "small", |
| 72 | "large" |
| 73 | ], |
| 74 | "merged": true, |
| 75 | "tokens": { |
| 76 | "input": 278415, |
| 77 | "output": 60000, |
| 78 | "cacheRead": 8538060, |
| 79 | "cacheWrite": 180000 |
| 80 | } |
| 81 | }, |
| 82 | { |
| 83 | "id": "fix-failing-check", |
| 84 | "kind": "revise", |
| 85 | "change": { |
| 86 | "files": 2, |
| 87 | "lines": 30, |
| 88 | "sensitive": [] |
| 89 | }, |
| 90 | "merged": false, |
| 91 | "tokens": { |
| 92 | "input": 19563, |
| 93 | "output": 11900, |
| 94 | "cacheRead": 599950, |
| 95 | "cacheWrite": 60860 |
| 96 | } |
| 97 | }, |
| 98 | { |
| 99 | "id": "what-does-this-flag-do", |
| 100 | "kind": "answer", |
| 101 | "merged": false, |
| 102 | "tokens": { |
| 103 | "input": 7560, |
| 104 | "output": 3600, |
| 105 | "cacheRead": 231840, |
| 106 | "cacheWrite": 40500 |
| 107 | } |
| 108 | }, |
| 109 | { |
| 110 | "id": "why-is-this-slow", |
| 111 | "kind": "answer", |
| 112 | "failsOn": [ |
| 113 | "small" |
| 114 | ], |
| 115 | "merged": false, |
| 116 | "tokens": { |
| 117 | "input": 19380, |
| 118 | "output": 6800, |
| 119 | "cacheRead": 594320, |
| 120 | "cacheWrite": 60500 |
| 121 | } |
| 122 | }, |
| 123 | { |
| 124 | "id": "review-small-change", |
| 125 | "kind": "review", |
| 126 | "change": { |
| 127 | "files": 3, |
| 128 | "lines": 80, |
| 129 | "sensitive": [] |
| 130 | }, |
| 131 | "merged": false, |
| 132 | "tokens": { |
| 133 | "input": 25626, |
| 134 | "output": 9000, |
| 135 | "cacheRead": 785864, |
| 136 | "cacheWrite": 68960 |
| 137 | } |
| 138 | }, |
| 139 | { |
| 140 | "id": "review-ci-change", |
| 141 | "kind": "review", |
| 142 | "change": { |
| 143 | "files": 2, |
| 144 | "lines": 40, |
| 145 | "sensitive": [ |
| 146 | "CI workflows" |
| 147 | ] |
| 148 | }, |
| 149 | "merged": false, |
| 150 | "tokens": { |
| 151 | "input": 21454, |
| 152 | "output": 8100, |
| 153 | "cacheRead": 657928, |
| 154 | "cacheWrite": 63480 |
| 155 | } |
| 156 | }, |
| 157 | { |
| 158 | "id": "review-medium-change", |
| 159 | "kind": "review", |
| 160 | "change": { |
| 161 | "files": 12, |
| 162 | "lines": 450, |
| 163 | "sensitive": [] |
| 164 | }, |
| 165 | "merged": false, |
| 166 | "tokens": { |
| 167 | "input": 65943, |
| 168 | "output": 15300, |
| 169 | "cacheRead": 2022252, |
| 170 | "cacheWrite": 108400 |
| 171 | } |
| 172 | }, |
| 173 | { |
| 174 | "id": "review-large-migration", |
| 175 | "kind": "review", |
| 176 | "change": { |
| 177 | "files": 80, |
| 178 | "lines": 4200, |
| 179 | "sensitive": [] |
| 180 | }, |
| 181 | "merged": false, |
| 182 | "tokens": { |
| 183 | "input": 355590, |
| 184 | "output": 36000, |
| 185 | "cacheRead": 10904760, |
| 186 | "cacheWrite": 180000 |
| 187 | } |
| 188 | }, |
| 189 | { |
| 190 | "id": "catch-up-with-conflict", |
| 191 | "kind": "update", |
| 192 | "merged": false, |
| 193 | "tokens": { |
| 194 | "input": 21583, |
| 195 | "output": 9000, |
| 196 | "cacheRead": 661903, |
| 197 | "cacheWrite": 63720 |
| 198 | } |
| 199 | }, |
| 200 | { |
| 201 | "id": "plan-checkout-redesign", |
| 202 | "kind": "plan", |
| 203 | "merged": false, |
| 204 | "tokens": { |
| 205 | "input": 54480, |
| 206 | "output": 28800, |
| 207 | "cacheRead": 1670720, |
| 208 | "cacheWrite": 98000 |
| 209 | } |
| 210 | } |
| 211 | ] |
| 212 | } |