Manual metadata CSV import #114

New issue

Open

opened 2026-07-28 06:06:05 +00:00 by claude-bot · 0 comments

claude-bot commented

2026-07-28 06:06:05 +00:00

Severity: MEDIUM

The problem

docs/circa-spec.md §8.4 defines a CSV format for user-supplied dates and notes that creates
versioned evidence. There is no issue and no code.

Anyone who scans thousands of photos keeps a spreadsheet. This is the cheapest bulk evidence source
in the entire design — one import can seed thousands of photos with human knowledge that would
otherwise be typed one photo at a time.

Worth confirming with the user: if a scanning-era spreadsheet already exists, this jumps from
medium to high priority.

Scope

Import per the spec's CSV format, matched to photos by filename or album plus sequence.
Create manual evidence rows with proper attribution and provenance to the import.
A dry-run mode reporting what would match, what would not, and why — before anything is written.
Idempotent re-import: re-running an amended sheet supersedes rather than duplicating.
Clear reporting of unmatched rows, which is the common failure and needs to be actionable.

Done when

A CSV can be imported, producing evidence with attribution
Dry-run reports matches and misses before writing
Re-import supersedes rather than duplicating
Unmatched rows are reported usefully

References

docs/circa-spec.md §8.4

## Severity: MEDIUM ## The problem `docs/circa-spec.md` §8.4 defines a CSV format for user-supplied dates and notes that creates versioned evidence. There is no issue and no code. Anyone who scans thousands of photos keeps a spreadsheet. This is the cheapest bulk evidence source in the entire design — one import can seed thousands of photos with human knowledge that would otherwise be typed one photo at a time. **Worth confirming with the user:** if a scanning-era spreadsheet already exists, this jumps from medium to high priority. ## Scope - Import per the spec's CSV format, matched to photos by filename or album plus sequence. - Create manual evidence rows with proper attribution and provenance to the import. - A dry-run mode reporting what would match, what would not, and why — before anything is written. - Idempotent re-import: re-running an amended sheet supersedes rather than duplicating. - Clear reporting of unmatched rows, which is the common failure and needs to be actionable. ## Done when - [ ] A CSV can be imported, producing evidence with attribution - [ ] Dry-run reports matches and misses before writing - [ ] Re-import supersedes rather than duplicating - [ ] Unmatched rows are reported usefully ## References - `docs/circa-spec.md` §8.4

claude-bot added this to the v0.4.0 milestone

2026-07-28 06:06:05 +00:00

claude-bot added the

backend

enhancement

labels

2026-07-28 06:06:05 +00:00