Bulk operations
Working with large numbers of emails
Sylvia can mark, archive, label, or delete hundreds of emails — but the precision of your command makes all the difference.
✓ Do this
mark emails from GitHub as unread
Filter by sender. Sylvia fetches only those emails — no risk of touching unrelated ones.
archive all LinkedIn emails from the last 30 days
A time or sender filter reduces the risk of touching the wrong emails.
mark Smile.CX emails older than 60 days as read
The more specific the criteria, the more predictable the result.
✗ Avoid this
mark all emails as unread
Without a number or filter, the operation can affect tens of thousands of emails.
try again
Without context, Sylvia may re-run the last command on a different subset of emails.
mark emails as unread in my inbox
The inbox only contains non-archived emails. To act on your whole mailbox, don't add "in inbox".
Where Sylvia looks
Inbox vs. your entire mailbox
Sylvia distinguishes between the inbox (non-archived emails) and the full mailbox (inbox + archive + sent). Knowing the difference avoids unexpected results.
Inbox only
Say "in my inbox" or "in inbox" to limit the search to non-archived emails. Useful when you only want to work with active, unarchived mail.
Entire mailbox
Without specifying a folder, Sylvia searches everything — inbox, archive, and sent. This is the correct behavior for "last N emails" or "all unread".
Trash and spam
Never included automatically. To act on those emails, say so explicitly: "emails in trash" or "emails in spam".
Sent emails
Sent emails are included in a generic search. If you want to exclude them, add a filter like "not from me" or specify the context more precisely.
Technical limits
Gmail API quota
Sylvia uses the official Gmail API. Like any tool that does, it is subject to Google's rate limits.
"Quota exceeded" — what it means
If you run many operations in quick succession in a single session (e.g. searching + marking hundreds of emails multiple times), Google may temporarily block further requests.
What to do: wait 30–60 minutes and try again. The limit resets automatically. It's not an account issue and requires no action in Gmail itself.
How to avoid it: run bulk operations one at a time, with a specific email count, and don't repeat the same command multiple times when something seems off.
What to do: wait 30–60 minutes and try again. The limit resets automatically. It's not an account issue and requires no action in Gmail itself.
How to avoid it: run bulk operations one at a time, with a specific email count, and don't repeat the same command multiple times when something seems off.
Conversation context
New session = clean memory
Sylvia tracks the context of the current conversation. Sometimes that helps; sometimes it gets in the way.
Sylvia says no emails found, but you can see them
If it previously replied "no unread emails found", it may use that as context. Open a new session with the + button and try again.
Context references work well
"archive that one", "reply to the last one", "the first in the list" — Sylvia resolves these from the results already shown in the session.
After a bulk operation, refresh Gmail
API changes don't always reflect instantly in the Gmail UI. After marking hundreds of emails, press F5 in Gmail to see the updates.
"Try again" can do different things
Without precise context, "try again" might re-run the command on a different subset. Better to rephrase the full command from scratch.
Recommended patterns
Commands that work well
These patterns produce predictable, reliable results.
Precise search
"show me unread emails from LinkedIn in the last 7 days" — criterion + sender + time range = targeted results.
Bulk labelling
"label the last 100 Jira emails as 'Work'" — fixed count + specific sender = safe, predictable operation.
Archive by sender
"archive all emails from noreply@github.com" — no risk of touching important emails, the filter is specific.
Sender analysis
"who emails me the most?" — Sylvia analyses your mailbox patterns and returns a ranked list of your top senders.