ai-memory vs deja-vu
Coding-agent memory tools compared
ai-memory and deja-vu help coding agents reuse context across sessions without requiring an LLM for their default search and recall paths. ai-memory captures activity into a git-backed Markdown wiki, while deja-vu indexes session history already stored on disk.
ai-memory: Share Long-Term Memory Across Coding Agents
A self-hosted Rust service that captures coding-agent activity into a searchable, git-backed Markdown wiki and hands work between agents. It suits developers and teams who want portable project memory without requiring LLM calls.
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.
| ai-memory | deja-vu | |
|---|---|---|
| Language | Rust | Go |
| License | MIT | MIT |
| Stars | 9.1k | 1.2k |
| Forks | 632 | 128 |
| Last analyzed | Oct 10, 2026 | Oct 19, 2026 |
Key differences
- ai-memory captures prompts, tool calls, and session boundaries through lifecycle hooks; deja-vu indexes existing agent session files, including JSONL and SQLite stores.
- ai-memory keeps Markdown pages as its source of truth and uses SQLite for search indexing; deja-vu builds an index from local session history.
- ai-memory is a Rust service that can run locally or as a shared server; deja-vu is a Go binary focused on local use.
- ai-memory supports multi-user authentication, attribution, and audit logs for shared servers; deja-vu offers local search, MCP recall, and hooks.
- ai-memory supports full-text search, entities, links, and optional vector search; deja-vu offers lexical search and optional semantic search through embedding services.
- Both projects are MIT-licensed and not archived; ai-memory is described as changing quickly, while deja-vu's project information also notes that most contributions come from one person.
Choose ai-memory if you…
- want a git-backed Markdown wiki that you can inspect, edit, and version-control.
- need a self-hosted shared server for project knowledge across machines or teammates, with personal handoffs kept separate.
- want activity captured through hooks and typed, claim-once handoffs between supported agents.
Choose deja-vu if you…
- want to search existing session history across Claude Code, Codex, Cursor, and other supported agents.
- need local CLI or terminal search with filters for agent, time, and project, plus commands for investigating past work.
- want hooks or MCP recall to surface prior decisions from indexed history, without first writing memory notes.
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.