DeepSeek Harness Plugin Hub
← Plugins

@dsh-external/dsh-engram-relay

Engram Relay

Large engram, small KV: an external engram adapter layer (cross-session hierarchical memory: global/project/session, AI-driven assignment decisions + causal link indexing + progressive disclosure + proactive awakening), implementing ultra-long-context memory sparse routing for DSH and coexisting wit

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

npx -y @deepseek-ai/dsh plugin --profile web add github:yjh051108/dsh-engram-relay#bcf9dc5b7219b06d1ab39190ada7a6c40c385123

Description

Large engram, small KV: an external engram adapter layer (cross-session hierarchical memory: global/project/session, AI-driven assignment decisions + causal link indexing + progressive disclosure + proactive awakening), implementing ultra-long-context memory sparse routing for DSH and coexisting with the official compact (the official compact frees KV; engram retains the details)

Compatibility and provenance

Engram Relay is published as @dsh-external/dsh-engram-relay and currently resolves to version 0.3.74. The Hub verifies its manifest and preserves the exact installation source for reproducible installs.

DSH compatibility
*
Runtime surfaces
web
Release source
github
Registry updated

Versions

0.3.74stable
0.3.12stable
0.2.0stable

Related plugins

More verified plugins in memory-context.

Hindsight Coding Agents@vectorize-io/hindsight-coding-agentsReflect-only Hindsight long-term memory for coding agents (harness-pluggable: opencode, …), with automatic background ingestion (no setup CLI).Memory Plugin@openviking/dsh-memory-pluginOpenViking memory and context bundle for DeepSeek HarnessMnemondsh-mnemonComposable three-tier memory control plane for DeepSeek Harness: persistent runtime context, searchable project documents, pluggable long-term memory, guarded strategies, WebUI, and headless tools.Dejadsh-dejaBrings the session history of twenty-one other coding agents into DeepSeek Harness: recall, session digest and per-file history tools over a local index, plus optional automatic recall before each step.