DeepSeek Harness Plugin Hub

发布与管理完整 Harness Profiles,发现适合你的插件。

探索

插件目录环境预设文档中心动态

社区

发布插件联系我们报告问题

相关链接

Plugin Hub GitHubDeepSeek Harness 官方项目系统状态隐私说明
© 2026 DeepSeek Harness Plugin HubPowered byPaxTech

独立、非官方社区项目,与 DeepSeek 官方无隶属、授权或背书关系。

DeepSeek Harness Plugin Hub
ProfilesPlugins分类动态文档登录管理 Profiles
ProfilesPlugins分类动态文档登录
← Plugins

dsh-dejavu-memory

Dejavu Memory

DeepSeek Harness 插件:DejaVu memory client——由您自己的 DejaVu MCP 服务器支持的自动化长期记忆(启动/读取/搜索/创建/更新/删除)。

插件会安装到这里;不确定时保持 web。

npx -y @deepseek-ai/dsh plugin --profile web add dsh-dejavu-memory@0.5.8
README兼容性版本

兼容性与来源证明

Dejavu Memory 以 dsh-dejavu-memory 发布,当前版本为 0.5.8。Plugin Hub 会校验它的 manifest,并保存精确安装来源,便于复现安装结果。

DSH 兼容范围
*
运行环境
any
发布来源
npm
Registry 更新时间
2026/11/4

版本

0.5.8stable
2026/10/30
0.5.7stable
2026/10/27
0.5.6stable
2026/10/26
查看其余 14 个版本收起版本
0.5.5stable
2026/10/22
0.5.4stable
2026/10/17
0.5.3stable
2026/10/16
0.5.2stable
2026/10/15
0.5.1stable
2026/10/15
0.5.0stable
2026/10/14
0.4.7stable
2026/10/1
0.4.6stable
2026/9/30
0.4.5stable
2026/9/29
0.4.4stable
2026/9/28
0.4.3stable
2026/9/24
0.4.2stable
2026/9/23
0.4.1stable
2026/9/18
0.4.0stable
2026/9/17

相关插件

正在加载相关插件…

最新版
0.5.8
DSH
*
HMR
重启进程
Tree shaking
未声明可安全裁剪
解包体积
30.1 kB
文件数
9
Surface
any
许可证
MIT
发布源
npm
GitHub
★ 0
周下载
356
安全扫描
✓ v0.5.8 扫描通过
最近提交
2026/11/3
查看源码 ↗项目主页 ↗
README Badge

点击下方 Badge 复制 Markdown,粘贴到 README 即可。

这是你的 Plugin?认领权益 · 优先安全扫描

验证 package.json 声明的 GitHub 仓库,即可管理这个公开页面。认领后,Hub 会优先安排当前版本的安全扫描,并在通过后公开展示结果。

认领这个 Plugin →
报告问题

相关插件

继续浏览 memory-context 分类下经过校验的插件。

Experimental Session Search@deepseek-ai/dsh-experimental-session-search为 Agents 提供搜索和读取当前工作区早期会话的工具。Hindsight Coding Agents@vectorize-io/hindsight-coding-agents面向编程代理的仅反思式 Hindsight 长期记忆(可插拔支持多种运行框架:opencode、……),自动在后台导入(无需设置 CLI)。Contextdsh-context用于上下文洞察和管理的 DeepSeek Harness 插件,提供上下文仪表板和上下文命令,帮助了解上下文的构成及其演变过程。Memory Plugin@openviking/dsh-memory-plugin适用于 DeepSeek Harness 的 OpenViking 记忆与上下文套件

README

dsh-dejavu-memory — long-term memory for DeepSeek Harness

dsh-dejavu-memory

Connects DeepSeek Harness to DejaVu: session-start boot + daily briefing, plus memory read / search / create / update / delete, backed by your own DejaVu MCP server on Cloudflare.

Port of pi-dejavu-memory — same protocol, same tool names.

The OpenAI/Codex plugin is the only client that opens the interactive memory graph. dsh uses the same server through tools only, with no graph UI and no user-facing MCP setup.

The OpenAI/Codex DejaVu Memory plugin showing the memory tree as an interactive graph

The OpenAI/Codex DejaVu Memory plugin inspector showing a node, its metadata, body text, and subtree

English · 中文

Tools

toolwhat it does
dejavu_bootload at session start: system://boot, system://recent/5, system://triggers, then best-effort system://briefing; afterward read system://focus (recent is a briefing subset — no need to re-read it after boot)
dejavu_readread a memory by URI (system://…, dejavu://agent, …)
dejavu_searchsearch memories (semantic + keyword / trigger recall via search_memory)
dejavu_createcreate a memory node ([Baseline]/[Deviation]/[Result]/[Reusable judgment])
dejavu_updatefull replace, patch (old_string/new_string), or append; optional relation
dejavu_deletedelete a memory by URI (delete_memory)

Quick start

dsh plugin --profile web add dsh-dejavu-memory

Requires your own DejaVu MCP server on Cloudflare.

- id: dejavu
  name: dsh-dejavu-memory
  config:
    mcp_url: https://dejavu.example.com/mcp
    mcp_auth: ""  # prefer mcp_headers for Access service token

For a server behind Cloudflare Access (e.g. dejavu.example.com), use the service token headers instead of mcp_auth:

- id: dejavu
  name: dsh-dejavu-memory
  config:
    mcp_url: https://dejavu.example.com/mcp
    mcp_headers:
      CF-Access-Client-Id: <your client id>
      CF-Access-Client-Secret: <your client secret>
keyrequiredmeaning
mcp_urlyesyour DejaVu MCP endpoint (Streamable HTTP)
mcp_authnolegacy; prefer mcp_headers for Cloudflare Access service token
mcp_headersnoextra headers merged into every MCP request (e.g. Cloudflare Access service token)

Tool naming

The dejavu_boot / dejavu_read prefixes make the memory tools explicit in the tool list and in the boot protocol, so agents call them directly instead of probing for a generic resource first.

License

MIT

Related

  • dejavu-memory for ChatGPT and Codex — same memory server, with the interactive graph UI
  • pi-dejavu-memory — same memory tools for Pi