[incognitojam]

Drizzle projects' pull requests and diffs listed drizzle/meta/_journal.json and the drizzle/meta/NNNN_snapshot.json files as source, so hundreds of lines of drizzle-kit bookkeeping sat above the code a reviewer needs to read and started expanded. TanStack Router's routeTree.gen.ts and other *.gen.* output had the same problem.

diffFileTier now classifies two more name patterns as generated:

• *.gen.*, alongside the existing *.generated.*.
• drizzle-kit's _journal.json and NNNN_snapshot.json, only when they sit directly in a meta/ directory.

The SQL migrations drizzle-kit writes stay in the source group, because they are what runs against the database. Repositories can still mark other generated paths through linguist-generated in .gitattributes. The classification is shared, so the change covers turn, working-tree, branch and pull request diffs on web and desktop, the changed-files card, and mobile review.

Validation

• New diffFileTier cases: drizzle journal and snapshot files and routeTree.gen.ts are generated; drizzle/0004_small_mentor.sql, src/meta/settings.json and src/codegen.ts stay source. vp test run packages/shared/src/diffFileOrder.test.ts passes (25 tests).
• Ran orderUnloadedDiffFiles on file names from a real drizzle pull request. The SQL migrations, README.md and src/ files come first, then home.test.ts, then the drizzle meta files, pnpm-lock.yaml and routeTree.gen.ts with the generated tag.
• Not checked in a running client. The UI already labels and folds the generated group, and this change only adds file names to it.

---

Written by an agent (Claude Code, claude-opus-5-5).

[github-actions]

Upstream divergence

Both sides are compared with upstream at the in-order tip 0ecb25a8d4. Excludes .repos/ and pnpm-lock.yaml.

Measure	Base	This PR	Change
Upstream files the fork changes	1,505	1,504	−1
Lines changed in them	78,367	78,252	−115
Upstream files the fork removes	42	43	+1
Fork-only files	549	550	+1
Fork-only lines	101,106	101,385	+279

Upstream files that start to differ (1)

• apps/web/src/components/pullRequest/pullRequestFileOrder.logic.test.ts removed: +0 −164 from upstream

Upstream files whose difference changes (1)

File	Change	Now differs by
packages/shared/src/diffFileOrder.ts	+5	163 lines

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ The exact PR base did not have a successful artifact. Baseline uses the latest successful main measurement shown below.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	5.0 KiB	4.9 KiB	−23 B (−0.5%)	6.8 KiB	✅
Codex	Thread snapshot wire	3.8 KiB	3.8 KiB	0 B (0.0%)	4.9 KiB	✅
Codex	Live turn WebSocket wire	1.2 KiB	1.2 KiB	−23 B (−1.9%)	2.0 KiB	✅
Codex	Live turn WebSocket decoded	20.9 KiB	20.8 KiB	−41 B (−0.2%)	29.3 KiB	✅
Codex	Live turn messages	2	1	−1 (−50.0%)	8	✅
Claude	Total thread wire	5.0 KiB	5.0 KiB	0 B (0.0%)	6.8 KiB	✅
Claude	Thread snapshot wire	3.8 KiB	3.8 KiB	0 B (0.0%)	4.9 KiB	✅
Claude	Live turn WebSocket wire	1.2 KiB	1.2 KiB	0 B (0.0%)	2.0 KiB	✅
Claude	Live turn WebSocket decoded	21.2 KiB	21.2 KiB	0 B (0.0%)	29.3 KiB	✅
Claude	Live turn messages	2	2	0 (0.0%)	8	✅

Baseline: 93cd1b4 · PR result: 92f7863 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 108.5 KiB
• Claude decoded thread snapshot: 108.8 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.