See where a memory came from — and fix it.
Every memory records which agent or source wrote it, and when. Open one to inspect its provenance and entity links, then edit or delete what an agent got wrong.
Open a memory to see where it came from.
Open a memory to inspect its provenance — source agent, timestamp, and entity links — then correct what the model got wrong.
Source and timestamp, always
From the very first write, every memory is stamped with which agent or source created it and when. Open one and you can trace exactly where it came from and which entities it links to — no guessing about why a memory exists.
- Provenance on every memory — source agent and timestamp, always
- Entity links you can follow — see what a memory connects to
- Not a black box — the differentiator vs. opaque memory APIs
The change is yours, not the model's
When an agent gets something wrong, you fix it. Edit a memory in place or delete it outright — in a click. The correction is yours, and it's shared across your connected agents so they all work from the corrected memory.
- Edit or delete in a click — the change is yours, not the model's
- Shared across agents — corrections reach every connected client
More capabilities
Five capabilities, one shared memory. Here are the other four.
Put your agents' memory in the open.
Start free and watch the memory form — then correct it when it drifts.
No credit card required · Free plan available · Bring your own OpenAI key