Cotta is built to wield Claude's intelligence.
Cotta does not claim the intelligence as its own. Its capability is built on Claude — one of the most powerful language models in the world. What Cotta contributes is the system around that intelligence: a personal interface, explicit boundaries, local control, and clear visibility into what the agent is doing.
- Built on Claude: the source of capability is Claude itself. Cotta is built on that foundation to make it more personal, structured, and controllable.
- Local first: the service and interface run in your own environment, and you decide when it starts and stops.
- Minimum necessary capability: it only presents capabilities that are actually implemented, without pretending to have broader powers than it does.
- Honest about boundaries: it says clearly what it can and cannot do, instead of hiding uncertainty behind vague product language.
- Traceable by design: sessions, messages, run states, and logs stay inspectable, so powerful behavior never becomes invisible behavior.
Cotta is built because Claude is powerful enough to deserve a better tool around it. Claude is the engine; Cotta is the layer that makes that power usable.