llms.txt@docs git:20260922.dbe721aB

Immutable. This exact content never changes and is served at /api/v1/blob/f38361b5c3f841ef.

# deja-vu

> Local memory for coding agents, built from the session transcripts they already write to disk. deja indexes what Claude Code, Codex, Cursor, opencode, Gemini CLI, Zed and twenty-eight more agents wrote on this machine — including the months before it was installed — and answers from them over MCP or the command line. BM25 over the transcripts: no model, no embeddings, and no network path unless you ask for one.

Two things distinguish it from memory tools that record forward: it starts full, because the history is already on disk, and recall arrives without being asked — at session start, on each prompt, before an edit or a command, after a command fails, and on the way back from a compaction, where deja keeps the task, files and commands the summary drops. Credentials are stripped as the index is built.

Install is one command (`curl -fsSL https://raw.githubusercontent.com/vshulcz/deja-vu/main/install.sh | sh`, or `brew install deja-vu`), followed by `deja install --auto` to wire the agents found on the machine.

## Start here

- [Getting started](https://vshulcz.github.io/deja-vu/guide/getting-started.html): install, wire an agent, and the first query
- [Why agents forget](https://vshulcz.github.io/deja-vu/guide/forgetting.html): what ends with a session and what survives on disk
- [Does my agent remember previous conversations](https://vshulcz.github.io/deja-vu/guide/does-my-agent-remember.html): what Claude Code, Codex and Cursor keep between sessions, and what they do not
- [Where history lives](https://vshulcz.github.io/deja-vu/guide/where-sessions-are-stored.html): the conversation history file each of the thirty-four agents writes, and its format
- [Credentials in transcripts](https://vshulcz.github.io/deja-vu/guide/credentials-in-transcripts.html): coding agents record command output verbatim, so keys and connection strings sit in plaintext transcripts; how to find them with deja secrets and what to do
- [Session files on disk](https://vshulcz.github.io/deja-vu/guide/session-files-on-disk.html): how to delete Claude Code sessions, how large ~/.claude/projects gets, what is inside a session file, and what deleting one costs
- [Delete Cline task history](https://vshulcz.github.io/deja-vu/guide/delete-cline-task-history.html): distinguish the modern CLI session store from the VS Code extension task tree and remove only the conversation data you intend to delete
- [CLI reference](https://vshulcz.github.io/deja-vu/guide/commands.html): every command and flag
- [Agents & MCP](https://vshulcz.github.io/deja-vu/guide/agents.html): the deja MCP tool and its modes, and what each harness supports

## Problems it answers

- [The agent lost the context you had](https://vshulcz.github.io/deja-vu/guide/lost-context.html): after a crash, a clear, or a restart
- [The context window is full](https://vshulcz.github.io/deja-vu/guide/context-window-full.html): what compaction keeps (77% of decisions, 0.2% of commands, measured) and what to do instead
- [Resuming a session](https://vshulcz.github.io/deja-vu/guide/resume-a-session.html): find it across every agent, reopen it in the one that owns it
- [The agent repeats a mistake you already fixed](https://vshulcz.github.io/deja-vu/guide/repeated-mistakes.html): the same error again, in a repository that solved it before
- [Finding a session](https://vshulcz.github.io/deja-vu/guide/find-a-session.html): search chat history across every agent at once
- [Switching agents](https://vshulcz.github.io/deja-vu/guide/switching-agents.html): move between Claude Code, Codex and Cursor without re-explaining
- [After compaction](https://vshulcz.github.io/deja-vu/guide/after-compaction.html): recovering the commands a summary dropped, and the packet deja hands back automatically
- [Auditing an agent](https://vshulcz.github.io/deja-vu/guide/auditing-agents.html): review agent changes when the diff cannot say what was tried
- [Exporting a conversation](https://vshulcz.github.io/deja-vu/guide/export-conversations.html): markdown, JSON, or a shareable page
- [Across machines](https://vshulcz.github.io/deja-vu/guide/sync-across-machines.html): sync between machines with no server
- [What memory costs](https://vshulcz.github.io/deja-vu/guide/token-cost.html): the measured token cost of recall
- [Stats card to PNG](https://vshulcz.github.io/deja-vu/card/): turn the card `deja stats --card` writes into a PNG, in the browser

## 中文

- [快速开始](https://vshulcz.github.io/deja-vu/zh/guide/getting-started.html)：安装、接上智能体、第一次查询
- [智能体记得之前的对话吗](https://vshulcz.github.io/deja-vu/zh/guide/does-my-agent-remember.html)：各个工具跨会话留下了什么
- [为什么会忘事](https://vshulcz.github.io/deja-vu/zh/guide/forgetting.html)：会话结束时消失的是什么，磁盘上留下的是什么
- [历史文件在哪里](https://vshulcz.github.io/deja-vu/zh/guide/where-sessions-are-stored.html)：三十四个智能体各自把会话写在磁盘的哪里，用什么格式
- [会话文件怎么删](https://vshulcz.github.io/deja-vu/zh/guide/session-files-on-disk.html)：Claude Code 的会话文件怎么删、能占多大、删掉之后失去什么
- [找回被删掉的会话](https://vshulcz.github.io/deja-vu/zh/guide/recover-a-deleted-session.html)：claude --resume 里找不到那次会话时，四种原因里有三种不是删除，真正还能找回的是什么
- [找回解决问题的那次会话](https://vshulcz.github.io/deja-vu/zh/guide/find-a-session.html)：按记得的词在所有智能体里同时搜历史，然后继续那次对话
- [隐私](https://vshulcz.github.io/deja-vu/zh/guide/privacy.html)：索引什么、脱敏什么、什么永远不出这台机器，以及怎么控制留下什么

## Per-harness setup

- [Claude Code, Codex and the rest](https://vshulcz.github.io/deja-vu/guide/harnesses.html): the full support matrix
- [Claude Code](https://vshulcz.github.io/deja-vu/guide/memory-for-claude-code.html)
- [Codex](https://vshulcz.github.io/deja-vu/guide/memory-for-codex.html)
- [Cursor](https://vshulcz.github.io/deja-vu/guide/memory-for-cursor.html)
- [Copilot CLI](https://vshulcz.github.io/deja-vu/guide/memory-for-copilot.html)
- [opencode](https://vshulcz.github.io/deja-vu/guide/memory-for-opencode.html)
- [DeepSeek Harness](https://vshulcz.github.io/deja-vu/guide/memory-for-dsh.html)
- [Kimi Code](https://vshulcz.github.io/deja-vu/guide/memory-for-kimi.html)
- [Zed](https://vshulcz.github.io/deja-vu/guide/memory-for-zed.html)
- [VS Code Copilot Chat](https://vshulcz.github.io/deja-vu/guide/memory-for-copilot-chat.html)
- [Grok Build](https://vshulcz.github.io/deja-vu/guide/memory-for-grok.html)
- [Gemini CLI](https://vshulcz.github.io/deja-vu/guide/memory-for-gemini.html)
- [Qwen Code](https://vshulcz.github.io/deja-vu/guide/memory-for-qwen.html)
- [Roo Code](https://vshulcz.github.io/deja-vu/guide/memory-for-roo.html)
- [prime-agent (PrimeIntellect)](https://vshulcz.github.io/deja-vu/guide/memory-for-prime.html)
- [Continue](https://vshulcz.github.io/deja-vu/guide/memory-for-continue.html)
- [Crush](https://vshulcz.github.io/deja-vu/guide/memory-for-crush.html)
- [OpenClaw](https://vshulcz.github.io/deja-vu/guide/memory-for-openclaw.html)
- [Goose](https://vshulcz.github.io/deja-vu/guide/memory-for-goose.html)
- [Cline](https://vshulcz.github.io/deja-vu/guide/memory-for-cline.html)
- [pi](https://vshulcz.github.io/deja-vu/guide/memory-for-pi.html)
- [omp (Oh My Pi)](https://vshulcz.github.io/deja-vu/guide/memory-for-omp.html)
- [Amp](https://vshulcz.github.io/deja-vu/guide/memory-for-amp.html): plugin API modelled on pi's, plus MCP
- [Hermes](https://vshulcz.github.io/deja-vu/guide/memory-for-hermes.html)
- [aider](https://vshulcz.github.io/deja-vu/guide/memory-for-aider.html): no MCP and no hooks, recall through the read-only context file
- [Antigravity](https://vshulcz.github.io/deja-vu/guide/memory-for-antigravity.html)
- [Senpi](https://vshulcz.github.io/deja-vu/guide/memory-for-senpi.html)
- [Kimchi Coding](https://vshulcz.github.io/deja-vu/guide/memory-for-kimchi.html)
- [gajae-code](https://vshulcz.github.io/deja-vu/guide/memory-for-gjc.html)
- [Command Code](https://vshulcz.github.io/deja-vu/guide/memory-for-commandcode.html)
- [ZCode](https://vshulcz.github.io/deja-vu/guide/memory-for-zcode.html)
- [CodeWhale](https://vshulcz.github.io/deja-vu/guide/memory-for-codewhale.html): read-only, nothing to install yet

## Evidence

- [Building deja into your own tool](https://github.com/vshulcz/deja-vu/blob/main/docs/INTEGRATING.md): the MCP server, the JSON surfaces, the hook commands, what is a contract and what is not, and how to isolate it in tests
- [Benchmarks](https://vshulcz.github.io/deja-vu/guide/benchmarks.html): 88.1% hit@1 on LongMemEval-S (470-question cleaned set), 70.6% on LoCoMo, both harnesses in the repository
- [How it compares](https://vshulcz.github.io/deja-vu/guide/compare.html): next to Mem0, Letta, memU, Memori, claude-mem and agentmemory
- [Day zero](https://vshulcz.github.io/deja-vu/guide/day-zero.html): eight tools installed on one machine that already held 19,195 sessions, then asked 100 questions only that history answers — build time, first answer, search latency, hit@1; the driver for every row is in the repository
- [Session format registry](https://vshulcz.github.io/deja-vu/registry/README.html): where each of the thirty-four agents stores its sessions, with fixtures
- [Privacy](https://vshulcz.github.io/deja-vu/guide/privacy.html): what is read, what is redacted, what never leaves

## Optional

- [Architecture](https://github.com/vshulcz/deja-vu/blob/main/docs/ARCHITECTURE.md)
- [Security model](https://github.com/vshulcz/deja-vu/blob/main/docs/SECURITY-MODEL.md)
- [Installing deja for an agent](https://github.com/vshulcz/deja-vu/blob/main/llms-install.md): the install page written for an agent doing it on a user's behalf
- [JSON output](https://github.com/vshulcz/deja-vu/blob/main/docs/json-output.md): the shape of every `--json` result
- [Compaction recovery](https://github.com/vshulcz/deja-vu/blob/main/docs/compaction.md): what is captured before a compaction, what is handed back, and the limits