Reviewed November 18, 2026 · By Claude-Mem
Find the thread without rereading every chat.
We connect recorded progress to the original request. Open a card to see what you asked for, the follow-ups and the latest result. Filter by project, search or time to get back to the work you care about.
Know where to put your attention.
My Work brings recorded requests for your input and results ready for review into one view. Open the evidence behind a card, make your decision and return to the agent conversation with a clear next move.
Keep the result within reach.
Reports, decks, images and other eligible local deliverables appear with the request that produced them. Our Stop hook can collect those files into a local deliverables bucket, keeping a copy even after the original scratch folder is gone.
Give the next session a useful starting point.
Copy handoff gathers the goal, recorded state, next step and recent updates. Bring it to the agent you choose, with the repository and files it needs. You carry the thread forward instead of reconstructing it from memory.
Questions before you start
How do I open My Work?
Install Claude-Mem, then open the worker URL printed by the installer, usually http://localhost:37777. My Work is the default local viewer. It reads your saved memory database without requiring a cloud account, a new model call or an internet connection. Memory & settings opens the wider saved history.
Does a status mean my agent is running or my result shipped?
In progress reflects the latest saved work state, not live agent presence. Needs you marks a recorded request for human input. Ready to check means the parent request was reported complete. Mark checked records your review of that version and offers Undo; it does not certify correctness or deploy a result. A substantive source update invalidates an older check.
Will my deliverable files follow me to another device?
The deliverables bucket stays on the computer that collected the files. Cloud Sync does not upload that bucket. Eligible files must be available to the local worker; remote-worker setups do not collect them. Copy handoff copies context as text and does not transfer files, assign work or launch an agent.
Does viewing local work upload my histories?
Opening the local viewer is separate from optional browser setup analysis and Cloud Sync. Browser setup lets you select tools or skip the scan. Choosing Discover My Work, Rules & Focus sends the selected prompt histories, with their session and project context, to cmem.ai for processing by a third-party AI model provider. Review that setup action before submitting.
Go deeper
Explore the source documentation and continue with the topic that matches your work.
Give your next session a head start.
Install Claude-Mem to capture your agents’ context and review the work it supports. Add CMEM Pro when you want to sync your agents’ memory across devices.

