Distributed databases are stepping out of the server and into your browser and edge. POST 1 demonstrates browser-based distributed DB experiments [1], while POST 2 shows vector-store-backed tooling and on-demand loading at the edge [2].
Browser-first DB experiments — POST 1 demonstrates browser-based distributed DB experiments. It spotlights the browser as a playground for distributed data ideas, expanding where data storage and processing can happen.
Edge tooling and cost savings — POST 2 centers on One-MCP, a FastAPI-based MCP server that acts as a semantic index and dynamic loader for tools [2]. Instead of loading every tool upfront, tools live in an external vector store and are retrieved on demand, so LLMs see only the most relevant tools and token budgets shrink [2]. Example curl flows show uploading tools and performing semantic search [2].
Closing thought: Taken together, these experiments hint at a future where latency, offline capability, and cost are weighed as compute and data move closer to users, not just into the cloud.
References
We Put a Distributed Database in the Browser – and Made a Game of It
Explores running a distributed database in the browser and gamifying it
View sourceShow HN: One-MCP: Unlimited tools MCP server without context bloat
Open-source MCP server enabling semantic tool search and on-demand loading from a vector store to reduce context size and costs
View source