improve-seed-packet-extraction

ARCHIVED

The current seed organizer prototype can scan seed packets, but extraction quality is hard to reason about because OCR, AI prompting, side-specific merging, normalization, storage mapping, and form prefill are tightly coupled. This change creates a more testable seed packet metadata pipeline so scanning techniques can evolve independently, users can still enter data manually when AI is unavailable, and accuracy can be measured against known-good packet data.

STAGES

  1. proposal05-13
  2. requirements05-13
  3. design05-13predates the Figma as-is + proposed pair
  4. tasks05-13
  5. build—
  6. archived—

CAPABILITY

  • seed-packet-metadata-extraction6 requirements

OUTCOMES — 28 OF 28 DONE, AND NOT MAPPABLE TO A CAPABILITY

This change organises its tasks by workstream rather than as spec scenarios, so progress cannot be split per capability.

WHAT HAPPENED

  1. TASKS

    proposal and specs and design and tasks landed

    2026-05-13
    • test(sso): cover extraction fallback verification
    • docs(openspec): finalize extraction archive

History read from a manifest built at 2026-09-30 (09e7bcc).

0 pull requests, found by commits touching the change folder · 2 commit(s) carry no pull request number