csyangwen/dsh-memory-evolve
A purely plugin-based implementation that brings "cross-session long-term memory + background self-evolution" capabilities to DeepSeek Harness: five-track memory · git branch awareness · in-turn self-review · skill self-evolution and skill manager · four-track to-do · COI scheduling · session broadcast · session search · prompt manager · temporary info notes—zero core modifications, zero runtime dependencies, ready to use upon installation and completely removable upon uninstallation.
Install
via GitHub · GitHub source
Installing from GitHub runs the project's build script, which pnpm blocks until you allowlist it — run the command once and pnpm prints the exact key to add under `allowBuilds` in ~/.dsh/profiles/web/pnpm-workspace.yaml.
The desktop app keeps its own profile, desktop, and only the dsh command it ships may change it — the npm-installed dsh refuses that profile. Install the command from the app menu (Manage dsh Command…), open Desktop once, quit it fully, run the line below, then reopen Desktop.
via GitHub · GitHub source
Installing from GitHub runs the project's build script, which pnpm blocks until you allowlist it — run the command once and pnpm prints the exact key to add under `allowBuilds` in ~/.dsh/profiles/desktop/pnpm-workspace.yaml.
Our install check ran in a web profile on npm's dsh, not in the desktop app.
What happened when we ran it
Installed cleanly when we ran it
Every command here is run in a throwaway container against a clean profile, and the result is whatever the harness recorded — not a guess from the source. Last run 8d agoon DeepSeek Harness 0.2.2-rc.1.
Show it in your README
For maintainers: the badge serves this listing's latest sandbox verdict, so a re-run updates it on its own — and it links readers to the full result here.
Due diligence
Security
No build script was blocked in our run.
Source: our sandbox install on DeepSeek Harness 0.2.2-rc.1, 8d ago.
package.json declares no preinstall, install or postinstall script.
Source: package.json on GitHub, read at our last sync. prepare is not recorded; it runs only on a GitHub install, and shows up under build scripts when pnpm blocks it.
shell commands (bash, exec, spawn …)
Source: a keyword match on the README — a hint about what the plugin touches, not an analysis of its code.
Not an audit. These are the record of one install and two text scans; nobody has read this plugin's code on your behalf. Plugins run with your agent's permissions.
What it is — The plugin provides cross-session long-term memory and background self-evolution support for DeepSeek Harness.
Who it is for — Developers switching sessions to maintain multiple projects suit this plugin, as the AI remembers project key facts and todos. If you seek tools without shell execution, this plugin requires executing shell commands.
Watch out — When installing from source, you need to manually allow the build script first. The sandbox test passed: installed in a new profile and registered into the profile by harness. No obvious pitfalls found.
The verdict — I would install it if cross-session memory is needed, since it is completely removable with zero core modifications.
Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.
dsh-memory-evolve is a DeepSeek Harness plugin that adds cross-session memory, task tracking, session coordination and background self-review without modifying the core or adding runtime dependencies. DeepSeek Harness loads it through the package’s cordis.patch.yml, which declares a dsh.bundle.patch; installation registers the plugin on the host automatically.
The plugin reads project context such as the working directory and Git branch, then writes persistent user, global, project and daily memory, project logs, four-track to-dos, skills, prompts and temporary notes. It can inject selected information into later turns, search local files, create and coordinate sessions, broadcast messages, dispatch work to external agents, and synchronise selected memory tracks through Git repositories. In-turn review is configured with reviewEnabled and reviewInterval; review is disabled by default. Many capabilities are also disabled until enabled in the plugin settings.
dsh-memory-evolve is intended for developers who work across projects or sessions and want persistent context, structured follow-up work or multiple AI sessions. It is a poor fit when persistent project data, session orchestration or external-agent dispatch is not wanted. The wake-up chain for idle sessions requires the main DeepSeek Harness process to remain running. The listed terminal-surface risk reflects access to terminal-facing agent workflows and local project files; installation should therefore be limited to profiles whose permissions match that use.
DeepSeek Harness registers dsh-memory-evolve from the bundled cordis.patch.yml after dsh plugin add; no manual insert entry is needed. Profile configuration belongs in cordis.patch.yml and overrides the bundle entry by id:
| Key | Default | Effect |
|---|---|---|
reviewEnabled |
false |
Enables in-turn memory review |
reviewInterval |
10 |
Reviews memory after this many user turns when review is enabled |
Set disabled: true on the dsh-memory-evolve entry to stop loading the plugin temporarily. The README also describes settings for enabling individual capabilities, including memory synchronisation, but does not provide a complete key list for those settings.
DeepSeek Harness manages the package with dsh plugin. The documented removal command is dsh plugin --profile web remove dsh-memory-evolve; removing the plugin cleans up its effects. After adding or changing the plugin, restart dsh web for the change to take effect. Existing profiles that manually inserted the plugin must remove that old insert entry, or the duplicate loader ID can prevent startup.
At the end of turns, the plugin records progress in project and daily logs. Project data is separated by working directory, and logs can include the current Git branch. Important project memories require confirmation before they become active context. Archived memories stop being injected until restored. Optional synchronisation stores selected memory tracks in dedicated Git branches, with an optional shared memory repository separate from project code.
The plugin can create internal sessions, place them in collaboration rooms, send broadcasts and coordinate external agents. Results can be written to project and daily logs. The README does not specify a minimum DeepSeek Harness version, operating-system support, environment variables or a complete on-disk file layout.
Written from the project's own documentation and kept in sync with it. Where the two disagree, the source is authoritative — read the README on GitHub
Same category
vectorize-io
Hindsight, agent memory that learns: long-term project memory with auto recall and retain, knowledge pages, deep reflection, and per-repo memory banks.
No one-line install — this plugin lives inside a larger repository and publishes no npm package.
volcengine
OpenViking memory and context bundle for DeepSeek Harness: pre-step auto-recall and profile injection, session capture, `viking://` URI guarding, and recall/write memory tools backed by an OpenViking server.
Installed cleanly when we ran it
MemTensor
Local-first memory plugin for DeepSeek Harness: layered L1/L2/L3 memory with automatic capture, manual memos_* retrieval tools, and a built-in memory viewer.
No one-line install — this plugin lives inside a larger repository and publishes no npm package.
agentscope-ai
Connects DeepSeek Harness to ReMe's local-first, self-evolving personal knowledge base: automatically captures completed main-agent conversations as user-owned Markdown memory, searches conversations and source material through reme_search with BM25, optional embeddings, and wikilink expansion, and schedules daily memory consolidation.
No one-line install — this plugin lives inside a larger repository and publishes no npm package.
agentscope-ai
Connects DeepSeek Harness to ReMe's local-first, self-evolving personal knowledge base: automatically captures completed main-agent conversations as user-owned Markdown memory, searches conversations and source material through reme_search with BM25, optional embeddings, and wikilink expansion, and schedules daily memory consolidation.
No one-line install — this plugin lives inside a larger repository and publishes no npm package.
zilliztech
Shared Markdown memory for DSH and other coding agents, with automatic capture, pre-step context injection, searchable recall, and memory-to-skill self-evolution through a review panel.
No one-line install — this plugin lives inside a larger repository and publishes no npm package.
