engram

engram gives AI coding agents a persistent memory, automatically capturing project context, conversation history and code structures so agents don't need to be retaught each session.

Strengths: single binary with zero dependencies; fast, accurate retrieval via SQLite and FTS5; multiple integration modes (MCP, HTTP API, CLI, TUI) make it adaptable to any agent workflow.

Weaknesses & barriers: the project is relatively young with a growing community; users must be comfortable with the command line or API integration; memory strategy design is left to the user, with no pre‑optimized out‑of‑the‑box policies.

Cost & deployment: open-source and free at its core; runs as a standalone binary or via Docker; a cloud option exists but no pricing was confirmed in the available evidence. Best suited for developers who use AI coding assistants frequently and need cross‑session continuity. Not ideal for one‑off tasks or users who prefer fully managed, zero‑setup services.