Dropstone is a model-independent AI runtime centered on persistent cross-session memory, learning from mistakes and human corrections, creating reusable skills, and self-verifying before delivery. It spans code, chat, documents, spreadsheets, and execution environments, with an SDK for portable memory across apps or models. Long-running tasks still require permissions and checkpoints.
Features
- Persistent memory
- Learning from corrections
- Reusable skills
- Self-verification and model-independent SDK
Use Cases
- Build long-running agents
- Reuse operating knowledge
- Move memory across models
- Verify outputs automatically
FAQ
Dropstone is a model-independent AI runtime centered on persistent cross-session memory, learning from mistakes and human corrections, creating reusable skills, and self-verifying before delivery. It spans code, chat, documents, spreadsheets, and execution environments, with an SDK for portable memory across apps or models. Long-running tasks still require permissions and checkpoints. Core capabilities include: Persistent memory, Learning from corrections, Reusable skills.
Common scenarios include: Build long-running agents, Reuse operating knowledge, Move memory across models.
