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

memory-eternal

Memory Eternal

记忆核心（Memory Eternal）：自研的 DeepSeek Harness 记忆插件，不移植任何既有记忆框架——对话结束后自动沉淀知识卡到本地 Markdown Vault（自研去重、自研 CJK 检索、可 git 管理），设置页提供图形化知识库（统计 / 搜索 / 知识图谱 + 侧边栏一键弹窗），Agent 通过 memory_recall 工具按需召回历史上下文。零人工干预。

插件会安装到这里；不确定时保持 web。

npx -y @deepseek-ai/dsh plugin --profile web add memory-eternal@0.20.0
README兼容性版本

兼容性与来源证明

Memory Eternal 以 memory-eternal 发布，当前版本为 0.20.0。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
web
发布来源
npm
Registry 更新时间
2026/11/4

版本

0.20.0stable
2026/11/3
0.19.0stable
2026/10/30
0.18.2stable
2026/10/26
查看其余 100 个版本收起版本
0.18.1stable
2026/10/26
0.18.0stable
2026/10/26
0.17.0stable
2026/10/24
0.16.0stable
2026/10/24
0.15.0stable
2026/10/23
0.14.3stable
2026/10/23
0.14.2stable
2026/10/23
0.14.1stable
2026/10/23
0.14.0stable
2026/10/23
0.13.0stable
2026/10/23
0.12.0stable
2026/10/22
0.11.0stable
2026/10/22
0.10.54stable
2026/10/20
0.10.53stable
2026/10/20
0.10.52stable
2026/10/19
0.10.51stable
2026/10/19
0.10.50stable
2026/10/19
0.10.49stable
2026/10/19

相关插件

正在加载相关插件…

最新版
0.20.0
DSH
*
HMR
重启进程
Tree shaking
未声明可安全裁剪
解包体积
2.4 MB
文件数
79
Surface
web
许可证
MIT
发布源
npm
GitHub
★ 131
周下载
2,383
安全扫描
✓ v0.20.0 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
0.10.48stable
2026/10/19
0.10.47stable
2026/10/19
0.10.46stable
2026/10/19
0.10.45stable
2026/10/19
0.10.44stable
2026/10/19
0.10.43stable
2026/10/19
0.10.42stable
2026/10/19
0.10.41stable
2026/10/19
0.10.40stable
2026/10/19
0.10.39stable
2026/10/17
0.10.38stable
2026/10/17
0.10.37stable
2026/10/17
0.10.36stable
2026/10/16
0.10.35stable
2026/10/16
0.10.34stable
2026/10/16
0.10.33stable
2026/10/16
0.10.32stable
2026/10/16
0.10.31stable
2026/10/15
0.10.29stable
2026/10/14
0.10.28stable
2026/10/14
0.10.27stable
2026/10/14
0.10.26stable
2026/10/13
0.10.25stable
2026/10/13
0.10.24stable
2026/10/13
0.10.23stable
2026/10/13
0.10.22stable
2026/10/13
0.10.21stable
2026/10/13
0.10.20stable
2026/10/13
0.10.19stable
2026/10/12
0.10.18stable
2026/10/12
0.10.17stable
2026/10/12
0.10.16stable
2026/10/9
0.10.15stable
2026/10/9
0.10.14stable
2026/10/9
0.10.13stable
2026/10/9
0.10.12stable
2026/10/9
0.10.11stable
2026/10/9
0.10.10stable
2026/10/9
0.10.9stable
2026/10/9
0.10.8stable
2026/10/8
0.10.7stable
2026/10/8
0.10.6stable
2026/10/7
0.10.5stable
2026/10/7
0.10.4stable
2026/10/7
0.10.3stable
2026/10/6
0.10.2stable
2026/10/5
0.10.0stable
2026/9/30
0.9.23stable
2026/9/30
0.9.21stable
2026/9/30
0.9.17stable
2026/9/30
0.9.16stable
2026/9/30
0.9.15stable
2026/9/29
0.9.14stable
2026/9/29
0.9.13stable
2026/9/28
0.9.12stable
2026/9/28
0.9.11stable
2026/9/28
0.9.10stable
2026/9/28
0.9.9stable
2026/9/28
0.9.8stable
2026/9/28
0.9.7stable
2026/9/28
0.9.6stable
2026/9/28
0.9.5stable
2026/9/28
0.9.4stable
2026/9/28
0.9.3stable
2026/9/14
0.9.2stable
2026/9/12
0.9.1stable
2026/9/12
0.9.0stable
2026/9/12
0.8.0stable
2026/9/5
0.3.9stable
2026/9/4
0.3.8stable
2026/9/4
0.3.7stable
2026/9/3
0.3.6stable
2026/9/3
0.3.5stable
2026/9/3
0.3.4stable
2026/9/3
0.3.3stable
2026/9/3
0.3.2stable
2026/9/3
0.3.1stable
2026/9/3
0.3.0stable
2026/9/3
0.2.0stable
2026/9/3
0.1.2stable
2026/9/3
0.1.1stable
2026/9/3
0.1.0stable
2026/8/31

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

🧠 memory-eternal — a second brain for your AI

English (default) · 简体中文

memory-eternal 20-second tour: review queue, token ledger, knowledge graph

▲ 20-second tour — review queue · token ledger · knowledge graph · live demo

Your AI forgets every session. Give it a memory it can trust.

Local-first, review-gated, measurable — one SQLite vault shared by DSH, Claude Code, Codex, Cursor and any MCP client.

  • 🔍 Review before it counts — new cards wait in a review queue; approve or reject from the Web UI or terminal; every change is written to an immutable audit log.
  • 🧾 Token ledger — see how much recall injected, how many distillation calls ran, and how often recall was actually useful.
  • 🏠 Local-first, no cloud — SQLite + Markdown on your own disk. No third-party memory framework, no account. Memory data never leaves the machine; only the News page fetches public RSS / Hacker News on demand (titles, excerpts and outbound links only).
  • ⭐ Core memory — promote approved cards to always-on memory; every session gets them automatically, and the review queue still decides what qualifies.
  • 🧭 Reader & layout — clickable table of contents in the card reader, plus a compact / normal / large layout setting for the whole memory UI.

⭐ If it saves you from explaining the same context again, a star helps a lot.
DSH one-liner: dsh plugin --profile desktop add memory-eternal

⚡ Quick start

DeepSeek Harness (desktop / web)

dsh plugin --profile desktop add memory-eternal
# reopen DSH → sidebar "Memory"

Claude Code, Codex or Cursor — same vault via MCP

npm i -g memory-eternal
dsh-memory connect claude    # or: codex / cursor

Any MCP client

npx -y -p memory-eternal dsh-memory mcp

Browser or terminal

dsh-memory open              # http://127.0.0.1:7999
npx -y -p memory-eternal dsh-memory tui   # live queue in the terminal

🧾 Does it actually save tokens?

Most memory products say they save tokens. We think you should check. The plugin ships a built-in memory ledger (Runtime status → Memory ledger) that records:

RecordedWhat it means
recall: hits + injected charactershow much memory actually entered the context
capture: LLM calls + input/output characterswhat distillation cost you
usefulness: useful / not-useful feedbackwhether recalled memories actually helped

The honest math:

net benefit ≈ avoided context cost − (injected characters + distillation calls × price)
ScenarioYou payYou avoid
Multi-session project with decisions and pitfalls~0.6–0.9k tokens per recallre-explaining context (~1.5–3k) or re-reading files (3–6k)
One-off question, low hit raterecall injectionusually nothing

Illustrative numbers, not a benchmark — measure your own with the ledger. If recall usefulness stays below ~60%, tune it down (fewer results, shorter summaries) or turn memory off.

🛡️ Review first, prove later

A memory system that silently ingests everything will eventually recall a wrong fact as truth. memory-eternal defaults to review-before-commit:

  • new cards land in a pending queue; nothing enters the main vault until a human approves it;
  • approve, reject or bulk-process from the Web UI, the terminal TUI, or the CLI;
  • every status change is appended to an immutable audit log (who / what / when / why) — enforced at the SQLite layer, not by convention;
  • agents can read and propose, but cannot approve their own cards;
  • deleted cards go to a recycle bin (30 days by default).

⭐ Core memory

Not everything should be injected into every session. Core memory is the small, explicit set of rules and preferences that must always win:

  • any approved card can be promoted from the reader or card list;
  • auto-capture may suggest core only for explicit long-term preferences / hard constraints, and core candidates always stay pending for review;
  • approved core cards are injected into the system prompt of every session — no memory_recall call needed;
  • set a card limit and character budget in Settings → Memory → Core memory;
  • in the knowledge graph, core nodes are orange diamonds with a ★;
  • never store secrets or API keys in a core card; keep only a pointer to the credential directory.

🎛 Reader & layout

  • click any heading in the reader's table of contents to smoothly scroll to that section;
  • Settings → Memory → Layout size offers compact / normal / large for the whole UI; it only affects display.

🧩 One vault, every agent

EntryWhat you get
DSH desktop / webembedded memory UI: review, graph, config, ledger
Claude Code / Codex / CursorMCP server, same local vault, dsh-memory connect
Any MCP clientnpx -y -p memory-eternal dsh-memory mcp
Terminaldsh-memory tui — live queue, select-all, batch approve/reject, long-form reader
Browserdsh-memory open → http://127.0.0.1:7999

📸 Screenshots

Knowledge graph: clusters by kind, tag edges Review inbox: type bar, queue and verdict pane Runtime status: database, backups, capture, version

▲ Knowledge graph · review inbox · runtime status

Native terminal audit: select-all, batch approve/reject Native terminal long-form reader

▲ Terminal-native review and long-form reading — dsh-memory tui

📚 Documentation & links

  • Full manual (English) · 完整手册（中文）
  • Private deployment (Docker / systemd)
  • Changelog · Architecture
  • Terminal plugin (dsh-tui) · Orca launcher · Cursor / VS Code extension
  • Report a bug

💬 Community

Questions, ideas, or DSH plugin talk — join the groups:

QQ groupWeChat group
QQ group QRWeChat group QR

Buy the author a coffee (optional):

WeChat PayWeChat Reward
WeChat Pay QRWeChat Reward QR

Sponsorship is optional; the plugin stays MIT and free. See SPONSOR.md.

📄 License

MIT — use it, self-host it, build on it.