Planned scope
- Requests users send to the assistant
- Threads they explicitly forward
- Conversations they include DTD in
- Relevant replies in those conversations
Security & privacy · beta design
Deliberate access.
Human control.
DTD is being designed around email conversations you intentionally include it in. During the closed beta, specific safeguards and capabilities are being developed and evaluated.
Product design goal
See only what you choose to share.
The intended email-first model works with requests, forwarded messages and conversations where DTD is deliberately included—rather than requiring blanket access to your mailbox.
Not a public feature yet
These are design intentions, not claims that every control is already implemented or externally audited. The task-handling inbox is not publicly available.
Beta participants will receive information relevant to the actual service before use.
Human oversight
Important decisions should stay with you.
We’re focusing on controls that let users change instructions, pause activity and require confirmation before consequential actions.
Clear authorizations
Routine follow-ups should be distinct from decisions that require explicit approval.
Stop and pause
Plain-language ways to end, defer or redirect work are part of the intended experience.
Credential handling
Any future AI provider credentials should be treated as secrets, with appropriate protective controls.
Transparent status
We will describe implemented protections as the beta matures, rather than claim certifications not yet achieved.
Mailing list vs product data
The mailing-list inbox is not the assistant.
Emailing addme@dotilldone.com is a request for beta updates and potential access. Please don’t send customer records, passwords, sensitive attachments or real tasks to this address. Read our mailing-list privacy notice →.
Closed beta · Join the list
Interested in the closed beta?
Join the mailing list to hear when more places are available.