Fix proofreading and snapshot races

Migrate runtime HTTP client to HTTPX2

Upgrade runtime dependencies

Preserve range identity in proofreading locators

Upload 9 files

Optimize literal sidecar construction

Upload 3 files

Add literal trigram sidecar index

Upload 6 files

Upload 6 files

Optimize search runtime performance

Optimize exact literal search verification

Add parse review confirmation

Keep accepted 160mb recovery limit

Probe 200mb snapshot recovery

Stream sidecar copy validation

Probe 160mb snapshot recovery

Test faster snapshot recovery limit

Parallelize snapshot restore work

Optimize serving paths and snapshot timing

Add generation-scoped first-screen bootstrap

Add editable parsing workflow

Refactor API modules and optimize runtime image

Validate proofreading changes and merge cost

Fix large and moved proofreading diffs

Refresh snapshots and generation caches safely

Serialize bucket snapshot publication

Raise facet filter selection limit to 500

Add proofreading version guards and idempotent dispatches

Implement convergence plan CLI for startup indexing

Persist identities for archive-scoped snapshot publication

Close serving SQLite connections after every operation

Raise proofreading auto-merge diff limit to 500

Fix incremental reindex generation and SQLite locking

Upload 2 files

Upload indexer.py

Upload 2 files

Upload 2 files

Upload 6 files

Upload bucket_snapshot.py

Upload 5 files

Upload 3 files

Upload 2 files

Upload 2 files

Upload 7 files

Upload main.py

Delete app/app.js

Delete app/style.css

Upload 2 files

Upload main.py