Skip to content

v0.1.1 — bug fixes + tests

Choose a tag to compare

@csitte csitte released this 23 Jun 09:18
5e956e5

Fixed

  • Attachment detection in search — fetch threads with full format (not metadata, which omits payload.parts, so detection always returned false).
  • Inline images no longer counted as attachmentscollectAttachments filters on Content-Disposition / Content-ID / X-Attachment-Id.
  • sweep_snoozed could miss threads when >100 shared a due date (pagination loss).
  • modify_labels resolves human-readable names → label ids (STARRED, ToDo, MCP/Snoozed); unknown names created (add) or skipped (remove).
  • Body decoding uses base64url — content with -/_ no longer corrupted.
  • Security: override uuid to ^11.1.1 (GHSA-w5hq-g745-h8pq); esbuild + vitest 4.1.9 bump.

Added

  • vitest test suite.

Changed

  • Refactor toward testability (pure collectBodies/collectAttachments/parseMessage, injectable API client).