Bulk export for the monthly Drive archive

[tytremblay]

Retention plan is a monthly export of recorded conversations into the team's Google Drive. export-conversation <id> [out.json] handles exactly one conversation, so "export everything" is a script that enumerates conversations and loops — which doesn't exist.

Needs: a subcommand that walks every conversation and writes a dated batch. Same no-Slack-credentials constraint as the other CLI commands.

Worth deciding at the same time whether the batch is one file per conversation (easier to redact for a guardian access request) or one archive per month (easier to retain and delete).

[tytremblay]

Superseded in its current form by the design session recorded in docs/adr/0001-metadata-exports-and-encrypted-snapshots.md.

Drive will not hold readable recorded messages. The scheduled job writes two artifacts instead:

• an export — readable, content-blind: roster, findings, role/group/setting changes, coverage summary, as dated CSVs
• a snapshot — the whole database, encrypted to a key the host does not hold, for disaster recovery only

The bulk-export-of-conversations subcommand this issue asks for is therefore not being built. The guardian-access-request case it was motivated by is served on demand by export-conversation <id>, which keeps reading a minor’ s messages a deliberate, attributed act rather than a scheduled dump into a folder whose sharing hawk-mod does not control.

Leaving this open until the new export feature lands, at which point it closes as superseded.