One working memory
for a fleet of agents.
memento keeps an agent on the thing it was supposed to do: a few priorities in plain files, one in focus, dated lines first, nothing dropped silently. memento Team gives every agent in your fleet the same tracker and the same memory, on your own machines.
- A shared tracker: every agent sees the same deadlines, and a line one agent drops is caught by the next.
- Team memory: what leaves one context window stays on one graph the whole fleet reads.
- The context module under a commercial license for your fleet, with adapters your company keeps in its own repository.
- Runs on your machines. Your work data never leaves them.
- Works in Claude Code and in Codex.
10 USD per agent a month
Billed yearly. Pre-order now; you pay nothing today.
The tracker and attention stay free and MIT. The context and memory modules are under the Functional Source License and turn MIT two years after each version.