Spooling
CLI Reference
Back to spooling.ai

spooling sync

Parse session files from every detected provider into the local database.

spooling sync [--no-embed] [--provider PROVIDER]

Walks every connected provider on this laptop, finds new or changed session files since the last run, parses them into rows in the local database, and embeds the message content into pgvector for semantic search.

Flags

FlagDescription
--no-embedSkip the embedding step. Faster sync, but spooling search won't find the new sessions until you re-run with embeddings.
--provider PROVIDEROnly sync the named provider (claude-code, cursor, codex, copilot, windsurf, kiro, antigravity).

How "new or changed" works

Spool tracks each session file by path + size in a local synced_files table. A file is reprocessed when:

  • it didn't exist in synced_files before, or
  • its size changed (which catches appended messages on long sessions).

Modification time is intentionally ignored. Many providers re-touch files for non-content reasons.

What sync does NOT do

sync is local-only. It does not push to Spooling Cloud. Use spooling push for that. The two-step separation is deliberate: most teams want to ingest everything locally and then choose what to share.

Common patterns

Sync just the providers you care about:

spooling sync --provider claude-code
spooling sync --provider kiro

Skip embeddings on a fast laptop (e.g. CI):

spooling sync --no-embed

Continuous sync: see spooling watch.

On this page