engram vs deja-vu
Persistent coding-agent memory compared
engram and deja-vu help coding agents and developers recover project knowledge across sessions. engram saves curated memories to a local database, while deja-vu indexes existing agent session histories for search.
engram: Give AI Coding Agents Persistent Memory
Engram is a Go-based memory system that helps AI coding agents retain useful project knowledge between sessions. It stores memories locally in SQLite and makes them available through MCP, a CLI, an HTTP API, and a terminal UI.
deja-vu: Search Coding-Agent History for Reusable Memory
deja-vu is a Go-based local memory tool that indexes existing session history from Claude Code, Codex, Cursor and 39 more coding agents. It provides search, MCP recall and hooks without requiring an LLM, so agents can reuse earlier decisions and fixes.
| engram | deja-vu | |
|---|---|---|
| Language | Go | Go |
| License | MIT | MIT |
| Stars | 7k | 1.2k |
| Forks | 724 | 128 |
| Last analyzed | Oct 3, 2026 | Oct 19, 2026 |
Key differences
- engram stores selected project knowledge in SQLite with FTS5; deja-vu builds a searchable index from existing session histories.
- engram offers MCP, CLI, HTTP API, and terminal UI access; deja-vu provides a CLI and MCP server, with optional session context supplied by supported harnesses.
- engram supports Git Sync and optional Cloud replication; deja-vu indexes and searches locally by default and includes tools for moving an index between machines.
- engram focuses on curated decisions, conventions, and handoffs; deja-vu can search session details such as files opened, commands run, and edit spans.
- Both are Go projects under the MIT license; the provided figures list 7k stars and 724 forks for engram, compared with 1.1k stars and 109 forks for deja-vu.
Choose engram if you…
- want to save curated project knowledge for later sessions rather than index entire conversation histories.
- need a local SQLite memory store with MCP, CLI, HTTP API, or terminal UI access.
- want optional Git Sync or Cloud replication for sharing or copying project memories.
Choose deja-vu if you…
- want to search existing coding-agent session histories without requiring an LLM or embeddings.
- need to find prior commands, file discussions, or edit details as well as conversation text.
- want session context or agent-initiated recall through the MCP server and supported harness integrations.
Compare the numbers side by side, or add more projects
This comparison is generated with AI from the OSRepos analyses of both projects. Always check each project's repository and documentation before choosing.