DeepSeek Harness Plugin Hub

Publish and manage complete Harness Profiles. Discover Plugins for your next setup.

Explore

PluginsPresetsDocsNews

Community

Publish a pluginContactReport an issue

Resources

Plugin Hub on GitHubDeepSeek HarnessSystem statusPrivacy notice
© 2026 DeepSeek Harness Plugin HubPowered byPaxTech

Independent and unofficial. Not affiliated with, authorized by, or endorsed by DeepSeek.

Memory — DSH Plugin for DeepSeek Harness
← Plugins

@klarkxy/dsh-memory

Memory

Long-term Memory: remember vocabulary, preferences, and recent activity per scope, consolidated by Dream.

The plugin will be installed here. Keep web if you are unsure.

npx -y @deepseek-ai/dsh plugin --profile web add @klarkxy/dsh-memory@0.1.7
READMECompatibilityVersions

Compatibility and provenance

Memory is published as @klarkxy/dsh-memory and currently resolves to version 0.1.7. The Hub verifies its manifest and preserves the exact installation source for reproducible installs.

DSH compatibility
*
Runtime surfaces
web
Release source
npm
Registry updated
10/17/2026

Versions

0.1.7stable
10/1/2026
0.1.6stable
10/1/2026
0.1.5
stable
9/28/2026
Show 5 more versionsCollapse versions
0.1.4stable
9/25/2026
0.1.3stable
9/24/2026
0.1.2stable
9/23/2026
0.1.1stable
9/23/2026
0.1.0stable
9/23/2026
Latest
0.1.7
DSH
*
HMR
Process restart
Tree shaking
Safe tree shaking not declared
Unpacked size
516.8 kB
Files
17
Surface
web
License
SEE LICENSE IN LICENSE
Source
npm
GitHub
★ 13
Weekly downloads
70
Last push
10/16/2026
View source ↗Project homepage ↗
README badge

Click the badge to copy Markdown for your README.

Do you maintain this Plugin?Claim benefit · Priority security scan

Verify the GitHub repository declared in package.json to manage this listing. After you claim it, Hub will prioritize a security scan of the current version and publish the result when it passes.

Claim this Plugin →
Report an issue
DeepSeek Harness Plugin Hub
ProfilesPluginsCategoriesNewsDocsSign inManage Profiles
ProfilesPluginsCategoriesNewsDocsSign in

README

@klarkxy/dsh-memory

DeepSeek Harness keeps hold of how you talk and what you are working on: the vocabulary you use, the preferences you stated, project facts, decisions and recent activity. Dream observes your own messages and consolidates them into scoped records; the same store can also hold procedural lessons, but Self Improve owns their creation and injection. Novel canon stays in the authoritative worldbook files.

简体中文

Standalone DSH Web

Requires Node.js ≥22 and DSH 0.1.7-rc.2. On a standalone DSH Web host, install:

npm install @klarkxy/dsh-memory

Load it explicitly:

dsh plugin --profile web add @klarkxy/dsh-memory

No application-private packages and no Self Improve are required — Memory stands on its own. The bundle starts enabled, and merely opening its UI does not call a model. Use Plugins → Memory to manage records, prompt injection and the Dream observation/consolidation switch.

Plugin page settings

Plugins → Long-term Memory carries two optional model rows, and neither is required:

RowUsed for
Dream modelconsolidation
Observation modelcontext observation

A saved route is used for that feature's own call as an explicit model; an empty selection follows the live session model and then the host default chat model. Only these two features are affected, and a saved route selects a model — it is not evidence of connectivity.

Behavior

Records and scope

ctx.aiMemory implements the shared MemoryService. New vocabulary and activity records may carry subject, domain, term/topic key, aliases, host observation time, literal event-time text and activity status. Project scope comes from the native session cwd, never an arbitrary client path, and automatic observation without a cwd is skipped rather than promoted to global.

Activity has a default seven-day freshness bound. Expiry removes it from recall; it does not prove completion. A new explicit update supersedes the previous matching subject/domain/topic while preserving history, and stable vocabulary has no automatic activity expiry. Existing authoritative task records and project documents remain authoritative.

Observation

At turn end, Dream observes original human messages only. A new session contributes its latest message; later batches include at most four new messages, each capped at 2000 characters. Accepted output requires exact quoted evidence from real human events, and pure continuation acknowledgements do not trigger inference. Context observation runs its own model call under the same resolved route rules.

Idle consolidation

Idle Dream runs after roughly 15 minutes of inactivity, at least 24 hours after the previous attempt, with at least three changes. It never consolidates lessons, crosses kinds or identities, extends source expiry, refreshes observation dates, or confirms candidate sources. Only all-active sources produce active replacements; otherwise the result stays candidate. Old manual records without referenceable evidence are left unchanged rather than assigned invented provenance. Corrections, deletion and lifecycle cancellation invalidate stale results.

Recall and injection

Recall contains at most five active records, about 800 tokens including the wrapper; lessons are omitted. For a continuation request, recent in-scope activity can restore the task context. Candidates, revoked, rejected, deleted, superseded and expired records are not injected, and current instructions and permissions take precedence.

Limits and compatibility

Observation cursors are persisted across restart and deletion. The 512-session cursor bound fails closed for new sessions instead of evicting deletion protection. Disabling Dream stops observation and idle consolidation, not Memory storage or Self Improve, and disabling the plugin preserves stored data.

Existing records remain readable without synthetic metadata. Update the shared service and both learning packages together when testing these new fields; merging source does not publish an npm release.

Verification

pnpm --filter @klarkxy/dsh-memory typecheck
pnpm --filter @klarkxy/dsh-memory test
pnpm --filter @klarkxy/dsh-memory build

API and exports

The package has three entry points:

  • . — the Cordis plugin (name, inject, apply), the MemoryRuntime service class and the shared constants CHAT_EVENTS_SLOT, MEMORY_RPC_CHANNEL, defaultSettings and projectIdFromCwd. apply provides the runtime as ctx.aiMemory and registers the host RPC channel.
  • ./contracts — browser-safe types and constants. MemoryRecord, MemoryService, MemoryQuery, NewMemoryRecord and the other shared interfaces come from the frozen @klarkxy/dsh-plugin-kit contracts; this package adds MemorySettings, DreamPlan, MemoryStatus, the recall/injection bounds and the /dsh-memory channel name.
  • ./client — the settings UI bundle loaded by the host web client.

MemoryRuntime implements the frozen MemoryService surface — list, create, update, promoteToGlobal, remove and recall — and adds settings (status/readStatus, updateSettings), the candidate lifecycle (accept, reject, revoke), manual records (createManualRecord), turn hooks (handlePreStep, observeSession) and the Dream lifecycle (previewDream, runIdleDream, applyDream, cancelDream).

The host RPC channel is /dsh-memory with endpoints status, settings.update, records.list, records.create, records.update, records.remove, records.accept, records.reject, records.revoke and dream.run.

Design and limits · Publishing · License