tinyMem (see `README.md`) is a local OpenAI-compatible HTTP proxy that gives small LLMs deterministic working memory by externalizing code artifacts and states into an immutable ledger/State Map combo so they never lose track of what they wrote.
ai proxy persistence small llm tiny-models llm-memory ai-memory agentic-ai agentic-coding deterministic-state
-
Updated
Dec 28, 2025 - Go