hindsight vs MemMolt
Agent memory that learns, not just recalls: retain/recall/reflect API over Postgres, SOTA on LongMemEval. Self-host via Docker with UI; Python/TS clients, any LLM provider. — versus — Structured long-term memory over MCP: an enforced bucket-thread-memo hierarchy in one SQLite file, hybrid FTS5 + vector search fused with RRF, local embeddings.
Both serve agent long-term memory with hybrid retrieval; memmolt is one SQLite file behind MCP with an enforced hierarchy, hindsight a self-hosted service that consolidates and reflects.
| hindsight | MemMolt | |
|---|---|---|
| Stars | 24k | 4 |
| Forks | 1.8k | 0 |
| Language | Python | JavaScript |
| License | MIT | MIT |
| Last activity | 2 days ago | 5 months ago |
| Topics | memory, agents | memory |
| Curated connections | 12 | 7 |
hindsight — the curator's take
Pick it when you want a deployable memory *service* whose pitch is learning — agents that get better over time, not a transcript search. The LongMemEval lead was independently reproduced (Virginia Tech, Washington Post), which is more than most memory vendors offer, and the LLM side is pluggable down to Ollama/LM Studio for fully-local stacks. NOT an embedded library: you run a Docker service with Postgres and talk to it over HTTP — overkill for a single coding agent wanting session notes. The ™ and Hindsight Cloud signal a commercial trajectory; watch where the open/paid line lands.
MemMolt — the curator's take
The anti-sprawl memory play: a forced 3-level hierarchy the agent can't turn into a jungle, with ~10ms hybrid search from a single SQLite file and zero cloud calls. Young and tiny (4 stars) — the schema idea is worth studying even if you don't adopt it. Skip if you want auto-capture; this is deliberate, curated memory.