SQLite Web: gitea.db

id

Primary key. INTEGER NOT NULL

user_id

INTEGER

op_type

INTEGER

act_user_id

INTEGER

repo_id

INTEGER

comment_id

INTEGER

is_deleted

INTEGER NOT NULL (default 0

ref_name

TEXT

is_private

INTEGER NOT NULL (default 0

content

{"Commits":[{"Sha1":"24a652278ccf5b431fadfe89021f9f76323f6e73","Message":"docs(evidence): rebind C48 follow-up evidence to d867d6e / 84ea344 (tsc-input-only cache key)\n\nReceipt gains a `revision` block: the verifier's finding, the v2 digest,\nsix consecutive prisma:generate key values before (b837ecfe / 22c21596\nalternating with identical d.ts) and after (aee939da x6 across both\nmanifest names, equal to the main checkout's key), regenerate-then-\ntypecheck 12/12 cache hits after the first v2 run, re-run negatives\n(truncated index.d.ts exit 2 / 24 error TS; client removed exit 1 MISSING),\n22/22 mutations, and gates on clean d867d6e and 84ea344 (pnpm check 20\nsteps 350/350 with 16/16 reuse, quick 17/17, check:kernel 17 steps,\ngovernance:status 12 VERIFIED / 8 STALE). Round 1 entries are kept.\nreports/*.latest.json refreshed from the 84ea344 runs; this commit itself\nwas not re-run.\n\nCo-Authored-By: Claude Opus 5.5 \u003cnoreply@anthropic.com\u003e\n","AuthorEmail":"hillao@juhailaoluodeMacBook-Pro.local","AuthorName":"juhailaoluo pro","CommitterEmail":"hillao@juhailaoluodeMacBook-Pro.local","CommitterName":"juhailaoluo pro","Timestamp":"2026-10-22T23:41:32-07:00"},{"Sha1":"84ea344d3b874621170ceed9fe4326f19ab24d33","Message":"docs(governance): describe the typecheck cache key as tsc inputs only\n\nCLAUDE/AGENTS dynamic area, README, CHANGELOG, gate-profile-wiring and the\nexperience library now say the key covers every generated *.d.ts plus the\nmanifest's types/typings/typesVersions/exports fields, list what is\nexcluded and why (manifest name and schema copy follow whichever backend\ngenerated last; the schema copy's staleness is judged by the freshness\ngate), record the verifier's finding on 7887b76 (key flipping between\nb837ecfe... and 22c21596... with identical d.ts), and update counts to\ntest:scripts 350 and 22 mutations. Static evidence line rebound to clean\nd867d6e; the receipt is updated in the next commit.\n\nCo-Authored-By: Claude Opus 5.5 \u003cnoreply@anthropic.com\u003e\n","AuthorEmail":"hillao@juhailaoluodeMacBook-Pro.local","AuthorName":"juhailaoluo pro","CommitterEmail":"hillao@juhailaoluodeMacBook-Pro.local","CommitterName":"juhailaoluo pro","Timestamp":"2026-10-22T23:37:33-07:00"},{"Sha1":"d867d6e03c10c3fcaadb3c7db7697dc1145135c9","Message":"test(governance): typecheck cache key ignores generation order, tracks every type input\n\nNew test: simulating the other backend generating last (manifest name and\nschema-copy comments change, freshness still passes) and changing main /\nbrowser leave the key unchanged; a one-byte index.d.ts change, a deleted\nd.ts, an added runtime/library.d.ts or client.d.mts, an exports `types`\ncondition, typesVersions, a missing manifest and a non-JSON manifest each\nchange it (each on a fresh fixture). The existing digest test now changes\nthe manifest `types` entry instead of its `name`, since `name` is no longer\nan input by design.\n\ntest:scripts 349 -\u003e 350. Mutation run re-executed on all 17 earlier\nmutations plus 5 new digest ones (whole manifest, schema copy, no manifest\nfields, no typesVersions, .d.ts only): 22/22 red, restored byte-identical.\n\nCo-Authored-By: Claude Opus 5.5 \u003cnoreply@anthropic.com\u003e\n","AuthorEmail":"hillao@juhailaoluodeMacBook-Pro.local","AuthorName":"juhailaoluo pro","CommitterEmail":"hillao@juhailaoluodeMacBook-Pro.local","CommitterName":"juhailaoluo pro","Timestamp":"2026-10-22T23:33:21-07:00"},{"Sha1":"1bd23f90972a52bad9d5d6b5fb8819ec9152cfc8","Message":"fix(governance): key typecheck cache only on what tsc reads from the generated client\n\nIndependent verification of 7887b76 found PRISMA_CLIENT_TYPES_DIGEST\nflipping between two values (b837ecfe... / 22c21596...) with all five\ngenerated .d.ts files byte-identical: the digest also hashed the generated\npackage.json (its `name` embeds a hash of whichever backend's schema\ngenerated last) and the schema.prisma copy (the two backend schemas differ\nin comments), and `pnpm prisma:generate` does not fix which backend writes\nthe shared directory last. Eight typecheck tasks re-ran for nothing.\n\nThe digest (now prisma-client-types-v2) covers every *.d.ts / *.d.mts /\n*.d.cts under the generated directory recursively, plus only the\ntype-resolution fields of the generated package.json (types, typings,\ntypesVersions, exports; missing/invalid manifest hashed as a fixed marker).\nThe schema copy is dropped on purpose: tsc never reads it, and its staleness\nis already judged by the freshness gate against each backend's source\nschema before any key is computed.\n\nSix consecutive prisma:generate runs now give one key (aee939da...) across\nboth manifest names, equal to the key of the main checkout's client.\n\nCo-Authored-By: Claude Opus 5.5 \u003cnoreply@anthropic.com\u003e\n","AuthorEmail":"hillao@juhailaoluodeMacBook-Pro.local","AuthorName":"juhailaoluo pro","CommitterEmail":"hillao@juhailaoluodeMacBook-Pro.local","CommitterName":"juhailaoluo pro","Timestamp":"2026-10-22T23:31:40-07:00"},{"Sha1":"7887b762894c2fc3d21f80e769242e38ef2c6023","Message":"docs(evidence): bind C48 follow-up evidence to 8f363f6 / 6f38665\n\nAdd reports/c48-followups-2026-10-23.json with gzipped raw logs:\nacceptance-first failure, before/after standalone typecheck (client\nremoved: 12/12 cached exit 0 -\u003e exit 1 MISSING without turbo; index.d.ts\ntruncated with schema untouched: 12/12 cached exit 0 -\u003e exit 2 with 24\nerror TS; stale schema -\u003e exit 1 STALE), real --reuse-current-run with three\ncontract-invalid sub-reports (each re-run alone, 15/16 reused), the\nself-test regression caught in the first profile draft, and 17 mutations.\n\nRefresh reports/*.latest.json from the clean 6f38665 runs: pnpm check\n20 steps 349/349 with 16/16 reuse, check:kernel 17 steps, governance:status\n12 VERIFIED / 8 STALE (same set as fcb264e). check:quick 17/17 was also run\n(isolated output, shared reports not written). This commit itself was not\nre-run.\n\nCo-Authored-By: Claude Opus 5.5 \u003cnoreply@anthropic.com\u003e\n","AuthorEmail":"hillao@juhailaoluodeMacBook-Pro.local","AuthorName":"juhailaoluo pro","CommitterEmail":"hillao@juhailaoluodeMacBook-Pro.local","CommitterName":"juhailaoluo pro","Timestamp":"2026-10-22T23:02:53-07:00"}],"HeadCommit":{"Sha1":"24a652278ccf5b431fadfe89021f9f76323f6e73","Message":"docs(evidence): rebind C48 follow-up evidence to d867d6e / 84ea344 (tsc-input-only cache key)\n\nReceipt gains a `revision` block: the verifier's finding, the v2 digest,\nsix consecutive prisma:generate key values before (b837ecfe / 22c21596\nalternating with identical d.ts) and after (aee939da x6 across both\nmanifest names, equal to the main checkout's key), regenerate-then-\ntypecheck 12/12 cache hits after the first v2 run, re-run negatives\n(truncated index.d.ts exit 2 / 24 error TS; client removed exit 1 MISSING),\n22/22 mutations, and gates on clean d867d6e and 84ea344 (pnpm check 20\nsteps 350/350 with 16/16 reuse, quick 17/17, check:kernel 17 steps,\ngovernance:status 12 VERIFIED / 8 STALE). Round 1 entries are kept.\nreports/*.latest.json refreshed from the 84ea344 runs; this commit itself\nwas not re-run.\n\nCo-Authored-By: Claude Opus 5.5 \u003cnoreply@anthropic.com\u003e\n","AuthorEmail":"hillao@juhailaoluodeMacBook-Pro.local","AuthorName":"juhailaoluo pro","CommitterEmail":"hillao@juhailaoluodeMacBook-Pro.local","CommitterName":"juhailaoluo pro","Timestamp":"2026-10-22T23:41:32-07:00"},"CompareURL":"luoanwu/base-framework/compare/05d26318c7ec4329875afd530b64839b00fd14d6...24a652278ccf5b431fadfe89021f9f76323f6e73","Len":8}

TEXT

created_unix

INTEGER

Update row 40779 in action