DynoTable

About DynoTable

DynoTable is a local-first desktop client for DynamoDB that adds the three things the AWS console does not have: real SQL, visual joins, and an AI agent that runs on your own credentials. The SQL Workbench accepts full SELECT statements with GROUP BY, COUNT, aggregations and multi-table joins over live DynamoDB data, which is a meaningful gap-filler for a database that has none of those natively. Smart Tables is the visual counterpart — draw relationships between tables on a canvas and work the joined view live, joining on the primary key or any global secondary index with the client picking the cheapest read, and with base tables never mutated because the view is a read-only projection. The AI agent runs through Amazon Bedrock on your own AWS credentials, which is the architecturally interesting part: there is no vendor server in the loop, the model executes inside your AWS account, you pay AWS at cost with no markup or per-seat AI fee, and prompts, schema and rows never cross your AWS boundary. The agent inspects real schema — keys, GSIs and item shapes — before querying so it prefers Query over Scan where it can, and any write is presented as a reviewable diff that nothing commits without approval. It supports AWS SSO, aws-vault, DynamoDB Local and PartiQL, exports to JSON, NDJSON and CSV, and ships an MCP server so Claude Code and Codex can read schema and items directly.

Key Features

DynoTable Pros & Cons

Who Is DynoTable Best For?

Tags