radimsem/remindb — An agentic memory database that cuts session tokens by 75–99%.

Description and use

radimsem/remindb is an open‑source, agentic memory database that stores an AI agent’s entire context in a single portable SQLite file, dramatically shrinking token usage by 75‑99% compared to traditional in‑memory approaches. By persisting the agent’s memory to disk, it enables fast, cost‑effective retrieval and seamless portability across environments, making it a lightweight backend solution for orchestrating long‑running or multi‑session AI workflows.

Problem

radimsem/remindb helps turn isolated prompts and tools into repeatable agent workflows.

Outcome

Early or unclear: treat as research material until maintenance, releases, docs, and issue activity are verified.

Workflow

• coordinate multi-agent workflows
• add tool-use pipelines
• standardize agent memory

Evidence

An agentic memory database that cuts session tokens by 75–99%. One portable SQLite file — your agent's memory, anywhere.

Source