|
72 | 72 | "r1a_production": 135 |
73 | 73 | } |
74 | 74 | ], |
75 | | - "successor_sha256": "8edca141d69cbb89d875d09c7618ce471f04ef41d1c0ecbc8520ed531344e037", |
| 75 | + "successor_sha256": "1615c02e5617b80b43d1d3dda57a4dd7a575cf659e6459a0ba946ba96179d4b5", |
76 | 76 | "grading_or_expected_rows_import": "forbidden" |
77 | 77 | } |
78 | 78 | }, |
|
110 | 110 | "semantic_change": "none", |
111 | 111 | "worker_pr_edit": "forbidden" |
112 | 112 | }, |
| 113 | + "review_correction": { |
| 114 | + "base_sha": "dce25ea306382b271391ce62036623996205c29e", |
| 115 | + "source_pr": 430, |
| 116 | + "worker_head_sha": "3a86a10b12122d6ff9bec70f5f62105157af25c8", |
| 117 | + "review_disposition": "requested_changes_preserved_open_unmerged", |
| 118 | + "superseded_patch_sha256": "38c0db5c2242a962b20fa2abd05c264fb08e36f6d9dc542fe5763ca69986c690", |
| 119 | + "selected_patch_sha256": "74776e524db39a8687963faa8a163dccd848fd9bf7ca81697ca3c86c5e5a0d5a", |
| 120 | + "selected_successor_paths": 21, |
| 121 | + "findings": [ |
| 122 | + "production publication preparation, not a test database adapter, owns complete session hierarchy materialization", |
| 123 | + "Q-WF-02 selects tool start and succeeded terminal boundaries independently across a half-open window", |
| 124 | + "the independent Q-WF-02 evaluator rejects duplicate call, tool, and state-change stable IDs", |
| 125 | + "Q-REV-03 delegation_metrics and token_deltas are direct fact-derived answers; exclusive/inclusive and scalar total delta formulas are explicit internal diagnostics because their result shapes cannot bind the frozen answer shapes" |
| 126 | + ], |
| 127 | + "worker_pr_edit": "forbidden", |
| 128 | + "implementation_acceptance": "not_granted" |
| 129 | + }, |
113 | 130 | "consumer_join": { |
114 | 131 | "query_compiler": { |
115 | 132 | "path": "src/codex_usage_tracker/agent_kernel/query/compiler.py", |
|
124 | 141 | }, |
125 | 142 | "synthetic_materialization": { |
126 | 143 | "paths": [ |
| 144 | + "src/codex_usage_tracker/agent_kernel/publication/preparation.py", |
| 145 | + "tests/agent_kernel/publication/test_preparation.py", |
127 | 146 | "tests/agent_kernel/fact_adapters/support.py", |
128 | 147 | "tests/agent_kernel/fixtures/oracles/cases_v2.py", |
| 148 | + "tests/agent_kernel/fixtures/oracles/database_replay.py", |
129 | 149 | "tests/agent_kernel/fixtures/tiny-v2/question-scenarios.json", |
130 | 150 | "tests/agent_kernel/fixtures/tiny-v2/oracle-bundle.json", |
131 | 151 | "tests/agent_kernel/fixtures/tiny-v2/manifest.json" |
132 | 152 | ], |
133 | 153 | "owner": "existing CK-08R1B worker", |
134 | 154 | "requirements": [ |
135 | | - "every selected session has one complete acyclic parent chain", |
| 155 | + "production publication materializes every selected session with one complete acyclic parent chain before query compilation", |
| 156 | + "the full 80-case database replay uses DatabaseV1FactCompiler rather than a test fact adapter", |
136 | 157 | "tool starts and terminals materialize their separately declared coordinates", |
137 | 158 | "canonical calls materialize exact measurement_mask and four token classes", |
138 | 159 | "Q-WF-02 session_selector excludes same-window foreign-session facts", |
|
148 | 169 | "production_role": "derive actual rows and compare only", |
149 | 170 | "copied_expected_rows": "forbidden", |
150 | 171 | "grading_source_imported_into_production": "forbidden" |
| 172 | + }, |
| 173 | + "q_rev_03_binding_decision": { |
| 174 | + "answer_fields": [ |
| 175 | + "delegation_metrics", |
| 176 | + "token_deltas" |
| 177 | + ], |
| 178 | + "classification": "direct_fact", |
| 179 | + "formula_diagnostics": [ |
| 180 | + "exclusive_inclusive_scope_v1", |
| 181 | + "side_by_side_delta_v1", |
| 182 | + "total_tokens_v1" |
| 183 | + ], |
| 184 | + "reason": "the frozen answer fields require a two-sided delegation object and a five-field token-delta object while the named formulas return one-side exclusive/inclusive and one scalar delta", |
| 185 | + "grading_or_expected_rows_as_input": "forbidden" |
151 | 186 | } |
152 | 187 | }, |
153 | 188 | "selected_successor_cohort": { |
154 | | - "preflight_base_sha": "97ea3aed8f67c7840a34b610e7e0588b7eaf3c4d", |
155 | | - "patch_sha256": "38c0db5c2242a962b20fa2abd05c264fb08e36f6d9dc542fe5763ca69986c690", |
| 189 | + "preflight_base_sha": "dce25ea306382b271391ce62036623996205c29e", |
| 190 | + "patch_sha256": "74776e524db39a8687963faa8a163dccd848fd9bf7ca81697ca3c86c5e5a0d5a", |
156 | 191 | "files": [ |
157 | | - {"path": "config/agent-kernel/logical-contract-v1.json", "predecessor_sha256": "27981ec9bac8c245306e02784b8730e8ff9216b33a81da9efb8409403431fe3d", "sha256": "3d7412156c735a9e1e497049ec86c8130c20632da27e009deed24d19c5836e84"}, |
158 | | - {"path": "config/agent-kernel/plan-operand-contract-v1.json", "predecessor_sha256": "19b9c69f50f67b01321e1f269cee2f70370ca64c0c666ff4dac2cc52efa314b4", "sha256": "0f21a484b66c36bb467bc659f2656fd20c31619e9bc7b6e8edd89250855edffe"}, |
159 | | - {"path": "experiments/physical-architecture/candidate_a/queries.py", "predecessor_sha256": "dd2a121480920cf96cd51ec43084a76a278009a8b72153b15963862d9619fe89", "sha256": "53c947fa8468db211428104aa4276e9138d6c48160de8c99821b5c509ae5929e"}, |
| 192 | + {"path": "config/agent-kernel/formula-contract-v1.json", "predecessor_sha256": "2e16599f244ac179313aa7bf7263caaddbd49f2c0984dbfa424ba89713e70ffa", "sha256": "3db5d4315d74f3c949cce6f6292f09026a5f26b4a1107c0434dbdd361c6a246c"}, |
| 193 | + {"path": "config/agent-kernel/logical-contract-v1.json", "predecessor_sha256": "27981ec9bac8c245306e02784b8730e8ff9216b33a81da9efb8409403431fe3d", "sha256": "cf10cd6efd7bb149bd41d92dc817e7656fe74b999639b129d019b4d69773ecd7"}, |
| 194 | + {"path": "config/agent-kernel/plan-operand-contract-v1.json", "predecessor_sha256": "19b9c69f50f67b01321e1f269cee2f70370ca64c0c666ff4dac2cc52efa314b4", "sha256": "e4a030893495e971e7890043c7ebe6c16843f5e457ad8c9aaa65202744364b06"}, |
| 195 | + {"path": "experiments/physical-architecture/candidate_a/queries.py", "predecessor_sha256": "dd2a121480920cf96cd51ec43084a76a278009a8b72153b15963862d9619fe89", "sha256": "4c14507bba3259d725a9432c2e193f7a5bdac14c880bdf145d548552aad01a7f"}, |
160 | 196 | {"path": "scripts/generate_ck07a_fixture.py", "predecessor_sha256": "d24be0a95b302a9c52d413c300f23536d8c0287512c4b446ed03f8d33ce0863f", "sha256": "f7adde83efb963121e841aec8d71ebd2e2be1fa3a1c2745d8e5ec05e6884cb68"}, |
161 | | - {"path": "src/codex_usage_tracker/agent_kernel/domain/plan_derivations_structural.py", "predecessor_sha256": "c63b5ae2528852768c7a0906c2dccd537587192116fd2cec956110bcc8ed1367", "sha256": "1c0441ca1d26c1d774aba9b53e0e10b04fcec2c6de7c1453a3e42aabcffa20ad"}, |
| 197 | + {"path": "src/codex_usage_tracker/agent_kernel/domain/plan_derivations_structural.py", "predecessor_sha256": "c63b5ae2528852768c7a0906c2dccd537587192116fd2cec956110bcc8ed1367", "sha256": "04673e50cdd7d3cc2c212295b64299b6a017bf86ebc74628e0c9bc1719aa0144"}, |
| 198 | + {"path": "src/codex_usage_tracker/agent_kernel/publication/preparation.py", "predecessor_sha256": "6689d61fbf6d7948e1958a9d0bc58b4ea326a7f04221914b74c0651e0be1e37c", "sha256": "d00df193b1bb44af6149734c67af9fe235dd7eed384c9dbb5c0732cf5a6c0868"}, |
162 | 199 | {"path": "src/codex_usage_tracker/agent_kernel/query/compiler.py", "predecessor_sha256": "0e57ddb03c5bff28eb74dc2eaa98f35491c649de71e7e9ba680c358a5adc0286", "sha256": "c3c43f044a43b5d2aed18cc03b88f1a5156a6570dd4d1cfc2f47a8b4d1833ffa"}, |
163 | | - {"path": "tests/agent_kernel/contracts/test_plan_derivations_structural.py", "predecessor_sha256": "853f5d6d44d17ef4a19bff0ab383d758d52541ec7c7b8bf92d008d93b3fb40fb", "sha256": "8531e34b4a1c11c4e55c49efbf369ee82126bdf2a4c373896b0142c6acdae915"}, |
164 | | - {"path": "tests/agent_kernel/contracts/test_plan_operand_contract.py", "predecessor_sha256": "72e9e205d56623eec9e26cba2fd8877190da1ee9675a779f49d527e989ad0988", "sha256": "b267512d0195c8e461a796af8c29671ce74ad71559ea6b6ffcac96269335ba6d"}, |
165 | | - {"path": "tests/agent_kernel/contracts/vectors/plan-operands-v1.json", "predecessor_sha256": "cc4d944354356136c45c6a22aa0a289849fe6c4804e4a5d5bd5574cac1ab72c2", "sha256": "8b83417dc5e7cd25ec1711fcd4b1152b90412d3594df918e3b5f2fdd363f845d"}, |
166 | | - {"path": "tests/agent_kernel/fact_adapters/database.py", "predecessor_sha256": "30c3f34d3847e75faeb5a5d543ec88b30939ad218ae38071e1929acddba6a63e", "sha256": "f81cad1e5c5875f9e6183047911265dbf21cb44266dfe0bfe1838cff8b26df92"}, |
| 200 | + {"path": "tests/agent_kernel/contracts/test_plan_derivations_structural.py", "predecessor_sha256": "853f5d6d44d17ef4a19bff0ab383d758d52541ec7c7b8bf92d008d93b3fb40fb", "sha256": "80a5df1cce9fa33e9203edd459d749d03530b4b76e5f4a7a2911129698972122"}, |
| 201 | + {"path": "tests/agent_kernel/contracts/test_plan_operand_contract.py", "predecessor_sha256": "72e9e205d56623eec9e26cba2fd8877190da1ee9675a779f49d527e989ad0988", "sha256": "1f7216e96232098d8efac7b19b205e399b9571d1e9ed6f3b489bb0c08f84b9d0"}, |
| 202 | + {"path": "tests/agent_kernel/contracts/vectors/plan-operands-v1.json", "predecessor_sha256": "cc4d944354356136c45c6a22aa0a289849fe6c4804e4a5d5bd5574cac1ab72c2", "sha256": "d23fb7a0b68f3b403f1add8d4b99a5a3d3d5881677f7bb2e144f4402cd2ee22b"}, |
| 203 | + {"path": "tests/agent_kernel/fact_adapters/database.py", "predecessor_sha256": "30c3f34d3847e75faeb5a5d543ec88b30939ad218ae38071e1929acddba6a63e", "sha256": "9ee0f23056aaa87a07981df56a0d130c5a976084fe945378182d9e0c9bc43471"}, |
167 | 204 | {"path": "tests/agent_kernel/fact_adapters/support.py", "predecessor_sha256": "2e8447613ce12fd0f82cb36af7ddebb2ba1e4eef1b6d53f31a990564b0be94c5", "sha256": "4e9ae26ce21909579bb5ccb1494cd55aaf1c2629ab547a50a8fda049829ad896"}, |
168 | | - {"path": "tests/agent_kernel/fact_adapters/test_contracts.py", "predecessor_sha256": "11f75a8025374455e90a0c4fb4eb9b2dc78bcce157e362fbe83aa001d6343536", "sha256": "cff89ea3b858df35695641cad6eed3ff2a04bf5dc9f2249b8d29108aa28d79b8"}, |
169 | | - {"path": "tests/agent_kernel/fixtures/independent/semantic.py", "predecessor_sha256": "8d46b9f65676087437386c270d975f2be8651a8eb299ee1e6d525b5eda913443", "sha256": "4c9729d9d0d7ad17f8ab3c341f5e71e377757bf5d16273aac8274f1efb9837d9"}, |
| 205 | + {"path": "tests/agent_kernel/fixtures/independent/semantic.py", "predecessor_sha256": "8d46b9f65676087437386c270d975f2be8651a8eb299ee1e6d525b5eda913443", "sha256": "1615c02e5617b80b43d1d3dda57a4dd7a575cf659e6459a0ba946ba96179d4b5"}, |
170 | 206 | {"path": "tests/agent_kernel/fixtures/oracles/cases_v2.py", "predecessor_sha256": "4ab32f4f924aef5f71e3b3d48478e5ae33eb86736a3b75de4b52621f4dae679b", "sha256": "7042fd002c1345c555a13148163cef326b3f6b4cf362ea7c922511e80c321b39"}, |
| 207 | + {"path": "tests/agent_kernel/fixtures/oracles/database_replay.py", "predecessor_sha256": "a22169f3884390ff7f98b871daf778aaae1fec7c5d0805eec3a9ce0db07219b6", "sha256": "73b3a0e796c0885d5ce09e8f5349f704a34138515b32729def3f139c17058467"}, |
171 | 208 | {"path": "tests/agent_kernel/fixtures/tiny-v2/manifest.json", "predecessor_sha256": "fb40a8a91d6ad537171e7a23e3f6fa9bd519080b513981b9483f9791e5e99e7d", "sha256": "4e33fc263f75cdd7d532d2972232b92c5320f58c0628154deaa74380d8875a8c"}, |
172 | 209 | {"path": "tests/agent_kernel/fixtures/tiny-v2/oracle-bundle.json", "predecessor_sha256": "97b78d11fa64eb2dac0a5f605bd9056f331ac9dd1fd210703e5f42b23935b40e", "sha256": "e5f682332b7529a85b420f1e1d2c543a5f867eef617fcb484ecc2d04c517da72"}, |
173 | 210 | {"path": "tests/agent_kernel/fixtures/tiny-v2/question-scenarios.json", "predecessor_sha256": "6ffca4917386c5bc13237952904d2a560a531e37c6eeba89b69ea53d76f35cd8", "sha256": "986d2f9689f0abf6515247655d75ad8c3587cf7635b201479801de0c94431014"}, |
174 | | - {"path": "tests/agent_kernel/test_ck08r1c_independent_evaluator.py", "predecessor_sha256": "db9c0dd58dbd24bb28a76cb1a603516686bf552e0303c451b5d1949c4b62e693", "sha256": "7fb178208c70c8e83525a0eb09018efa523766c2f8ac8aa8aecb097eb4e900c0"} |
| 211 | + {"path": "tests/agent_kernel/publication/test_preparation.py", "predecessor_sha256": "2b43cdfca0d935d4f08aa44a16bd176713ddf3ef7083ff3e76b42d4d0c8687e3", "sha256": "d0c277af4c655efb9b57c9bd7605592ce8a4959a12ffdad030935af5a4df6ce9"}, |
| 212 | + {"path": "tests/agent_kernel/test_ck08r1c_independent_evaluator.py", "predecessor_sha256": "db9c0dd58dbd24bb28a76cb1a603516686bf552e0303c451b5d1949c4b62e693", "sha256": "defac63d1ee27c83a67c4f3f9e8dae24d713a263582614b83c9f9b20279d4ce2"} |
175 | 213 | ], |
176 | 214 | "focused_test_identities": [ |
177 | 215 | { |
178 | 216 | "path": "tests/agent_kernel/contracts/test_plan_derivations_structural.py", |
179 | | - "sha256": "8531e34b4a1c11c4e55c49efbf369ee82126bdf2a4c373896b0142c6acdae915" |
| 217 | + "sha256": "80a5df1cce9fa33e9203edd459d749d03530b4b76e5f4a7a2911129698972122" |
| 218 | + }, |
| 219 | + { |
| 220 | + "path": "tests/agent_kernel/publication/test_preparation.py", |
| 221 | + "sha256": "d0c277af4c655efb9b57c9bd7605592ce8a4959a12ffdad030935af5a4df6ce9" |
180 | 222 | }, |
181 | 223 | { |
182 | 224 | "path": "tests/agent_kernel/test_ck08r1c_independent_evaluator.py", |
183 | | - "sha256": "7fb178208c70c8e83525a0eb09018efa523766c2f8ac8aa8aecb097eb4e900c0" |
| 225 | + "sha256": "defac63d1ee27c83a67c4f3f9e8dae24d713a263582614b83c9f9b20279d4ce2" |
| 226 | + }, |
| 227 | + { |
| 228 | + "path": "tests/agent_kernel/fixtures/oracles/database_replay.py", |
| 229 | + "sha256": "73b3a0e796c0885d5ce09e8f5349f704a34138515b32729def3f139c17058467" |
184 | 230 | } |
185 | 231 | ], |
186 | 232 | "focused_validation": { |
187 | | - "result": "182 passed", |
| 233 | + "result": "135 passed focused preflight", |
188 | 234 | "case_count": 80, |
189 | 235 | "independent_rows_equal_production_rows": true, |
190 | 236 | "independent_grades_equal_frozen_grades": true, |
191 | 237 | "fixture_source_jsonl_unchanged": true, |
192 | 238 | "review_mutations": [ |
193 | | - "Q-WF-02 no-match and same-window foreign-session selector isolation", |
194 | | - "Q-WF-02 valid before-window and after-window tool exclusion", |
195 | | - "Q-REV-03 missing, null, and mask-value mismatch while capability is unavailable" |
| 239 | + "Q-WF-02 start-before/terminal-inside and start-inside/terminal-after straddling", |
| 240 | + "production publication complete hierarchy plus dangling and cyclic rejection", |
| 241 | + "independent evaluator duplicate call, tool, and state-change stable-ID rejection", |
| 242 | + "Q-REV-03 direct fact answers and bound internal formula diagnostics" |
196 | 243 | ] |
197 | 244 | } |
198 | 245 | }, |
|
201 | 248 | "missing, null, malformed, or value-mismatched canonical-call measurement_mask rejects even when structural-context capability is unavailable", |
202 | 249 | "Q-WF-02 session selector excludes same-window foreign-session calls, tools, and state changes", |
203 | 250 | "valid Q-WF-02 tools before or after the selected half-open window are excluded", |
| 251 | + "Q-WF-02 start and succeeded terminal membership are selected independently for both straddling directions", |
204 | 252 | "partial or null start coordinate rejects", |
205 | 253 | "partial or null terminal coordinate rejects", |
206 | 254 | "terminal without start rejects", |
207 | 255 | "succeeded lifecycle without terminal succeeded transition rejects", |
208 | 256 | "terminal ordered before start rejects", |
209 | 257 | "duplicate stable session, turn, call, tool, state-change, or resource identity rejects", |
| 258 | + "Q-REV-03 declared formula diagnostics are internal and bound while answer fields remain exact fact-derived values", |
210 | 259 | "grading sentinel and grading inaccessible leave both evaluators unchanged", |
211 | 260 | "production-source mutation changes production only and exposes a mismatch", |
212 | 261 | "canonical-fact mutation changes both evaluators", |
|
251 | 300 | "tests/kernel/test_kernel_scope.py" |
252 | 301 | ], |
253 | 302 | "implementation_allowlist": [ |
254 | | - "the exact eighteen selected_successor_cohort paths", |
| 303 | + "the exact twenty-one selected_successor_cohort paths", |
255 | 304 | "focused tests for query compiler, fact adapters, structural derivation, R1C evaluator, fixture reconciliation, and fixture generation" |
256 | 305 | ], |
257 | 306 | "locks": [ |
|
0 commit comments