Skip to main content

ReMe Studio

English | 简体中文

ReMe Studio is the local web workspace for ReMe. It lets you browse and edit user-owned workspace files, explore memory links, and chat with the ReMe Agent without moving durable memory into a separate application database. Search indexes, graphs, and other derived metadata remain rebuildable from the source files.

ReMe Studio workspace

Installation

Install Studio together with ReMe's optional integrations:

pip install "reme-ai[core]"

For Studio without the other optional integrations, use pip install "reme-ai[web]". The base reme-ai package is headless and does not include the frontend assets.

Features

  • Workspace browsing: browse the full workspace or focus on journal and knowledge files through dedicated views. The navigator refreshes as files change on disk.
  • Markdown editing and preview: open multiple files in tabs, render Markdown front matter and GitHub Flavored Markdown, edit with Monaco, save with optimistic modification-time checks, and download files locally.
  • Memory graph: inspect indexed wikilinks under the wiki, personal, and procedure knowledge roots, follow inbound and outbound links, and open the corresponding Markdown source.
  • Agent chat: stream conversations with the read-only workspace Agent, see tool calls and token usage, and drag workspace files into the conversation as references.
  • Service management: inspect service and component memory usage, review the effective redacted configuration and version, and rebuild derived indexes without modifying source memory files.
  • Personalization: switch between English and Chinese, and use light, dark, or system appearance.

Requirements

  • Python 3.11 or newer with ReMe installed.
  • A running ReMe HTTP service. Agent chat additionally requires a working Agent and model configuration.
  • Node.js 22.13 or newer is required only when developing or building Studio from source.

See the repository README for ReMe installation and backend configuration.

Development

Start ReMe from the repository root, then run the frontend in another terminal:

# Terminal 1, from the repository root
reme start

# Terminal 2
cd website
npm install
npm run dev

Open http://localhost:3000. The frontend connects to http://127.0.0.1:2333 by default. Override it when needed:

NEXT_PUBLIC_REME_API_URL=http://127.0.0.1:8000 npm run dev

ReMe-hosted static build

ReMe can serve Studio from the same FastAPI process as its HTTP API. Build the static variant and restart ReMe:

cd website
npm ci
npm run build:static
cd ..
reme start

Open http://127.0.0.1:2333. The static build uses same-origin requests by default. For standalone static development, run npm run dev:static and set VITE_REME_API_URL to the running ReMe service URL when necessary.

The regular npm run build command remains the vinext/Sites deployment build; npm run build:static creates dist-static/ exclusively for FastAPI and Python package distribution.

Configuration

The workspace hides dotfiles and dot-directories. It displays only Markdown and text files by default. Configure the allowed extensions as a comma-separated list in .env.local:

NEXT_PUBLIC_REME_WORKSPACE_EXTENSIONS=md,txt,mdx

The memory graph requires an index built by ReMe. Rebuilding the index from the Studio settings regenerates derived data from workspace files and does not modify the source memory.

Checks

npm run format:check
npm run lint
npm run build
npm run build:static
npm test

简体中文

English | 简体中文

ReMe Studio 是 ReMe 的本地 Web 工作区。你可以在这里浏览和编辑自己拥有的工作区文件、探索记忆之间的联系,并与 ReMe Agent 对话,而无需将持久记忆迁移到独立的应用数据库中。搜索索引、图谱和其他派生元数据均可根据源文件重建。

ReMe Studio 工作区

安装

安装 Studio 和 ReMe 的可选集成功能:

pip install "reme-ai[core]"

如果只需要 Studio,不需要其他可选集成,可以使用 pip install "reme-ai[web]"。基础 reme-ai 包以无界面模式分发, 不包含前端资源。

功能

  • 浏览工作区:浏览完整工作区,或通过独立视图聚焦日记和知识文件;磁盘中的文件发生变化后,导航器会自动刷新。
  • Markdown 编辑与预览:在多个标签页中打开文件,渲染 Markdown front matter 和 GitHub Flavored Markdown,使用 Monaco 编辑器编辑,通过修改时间检查安全保存,并可将文件下载到本地。
  • 记忆图谱:查看知识库 wikipersonalprocedure 目录中已索引的 wikilink,检查入链和出链,并从图谱打开对应的 Markdown 源文件。
  • Agent 对话:与只读工作区 Agent 进行流式对话,查看工具调用和 token 用量,还可将工作区文件拖入对话作为引用。
  • 服务管理:查看服务及组件的内存使用情况、当前生效的脱敏配置和版本,并在不修改记忆源文件的情况下重建派生索引。
  • 个性化设置:切换中英文界面,并使用浅色、深色或跟随系统的外观。

环境要求

  • Python 3.11 或更高版本,并已安装 ReMe。
  • 正在运行的 ReMe HTTP 服务。Agent 对话还需要可用的 Agent 和模型配置。
  • 只有从源码开发或构建 Studio 时才需要 Node.js 22.13 或更高版本。

ReMe 的安装和后端配置请参阅仓库中文 README

本地开发

先在仓库根目录启动 ReMe,然后在另一个终端运行前端:

# 终端 1:仓库根目录
reme start

# 终端 2
cd website
npm install
npm run dev

打开 http://localhost:3000。前端默认连接 http://127.0.0.1:2333,需要时可覆盖该地址:

NEXT_PUBLIC_REME_API_URL=http://127.0.0.1:8000 npm run dev

由 ReMe 托管的静态构建

ReMe 可以通过提供 HTTP API 的同一个 FastAPI 进程托管 Studio。构建静态版本并重启 ReMe:

cd website
npm ci
npm run build:static
cd ..
reme start

打开 http://127.0.0.1:2333。静态构建默认使用同源请求。进行独立的静态开发时,运行 npm run dev:static;如有需要,将 VITE_REME_API_URL 设置为正在运行的 ReMe 服务地址。

常规的 npm run build 命令仍用于 vinext/Sites 部署构建;npm run build:static 仅为 FastAPI 和 Python 包分发生成 dist-static/

配置

工作区会隐藏点文件和点目录,并且默认只显示 Markdown 和文本文件。可以在 .env.local 中通过逗号分隔的列表配置允许显示的扩展名:

NEXT_PUBLIC_REME_WORKSPACE_EXTENSIONS=md,txt,mdx

记忆图谱依赖 ReMe 构建的索引。在 Studio 设置中重建索引时,只会根据工作区文件重新生成派生数据,不会修改记忆源文件。

检查

npm run format:check
npm run lint
npm run build
npm run build:static
npm test

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

reme_ai_studio-0.4.1.7.tar.gz (3.6 MB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

reme_ai_studio-0.4.1.7-py3-none-any.whl (3.7 MB view details)

Uploaded Python 3

File details

Details for the file reme_ai_studio-0.4.1.7.tar.gz.

File metadata

  • Download URL: reme_ai_studio-0.4.1.7.tar.gz
  • Upload date:
  • Size: 3.6 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for reme_ai_studio-0.4.1.7.tar.gz
Algorithm Hash digest
SHA256 0c0d5626ec49fee355f0a855eda6bb5a53eedae5a9f97f7317385739312d8e69
MD5 9f6fc180fe036e855036d57f67ba1f68
BLAKE2b-256 4ea7689f04a4522c07655fda3e9d87f22c35438afbd33fb62c900ab48c8b5666

See more details on using hashes here.

File details

Details for the file reme_ai_studio-0.4.1.7-py3-none-any.whl.

File metadata

File hashes

Hashes for reme_ai_studio-0.4.1.7-py3-none-any.whl
Algorithm Hash digest
SHA256 d100b55c8fc890a41e5c803e357bbbd98a9e6561fb91379b5f09427acf760c0e
MD5 249edc385909e194a1c7c397a4394186
BLAKE2b-256 797029dd9a1a634a4612915990b122377c73edae458d08b508b6e4370bd60053

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page