TL;DR
Memory-assembly team for DeepSeek Harness: every message triggers a five-step cognition loop (judge, retrieve, assemble, feedback, calibrate) carried by role-playing subagents (judge, retriever, archivist, advocate, editor, scribe), using BM25 retrieval, time-ordered material, and LLM assembly to build useful material packs injected into linghun. Works with dsh-linghun.
Install in the web profile
package.json
Official DSH compatibility Official version baseline recorded 2026-11-05
Checks apply to the package versions below. Installation, startup and functionality are separate checks.
DSH 0.2.2-rc.1 (latest) Version check pending
DSH 0.2.2-rc.2 (next) Version check pending
Official release history ↗ dsh plugin --profile web add github:syyr1987/dsh-linghun-assemblerCopy
Declared command, not sandbox-tested.
Install verification
Pending
Tested official runtime Exact version not recorded — recheck needed
DSH engine Not declared — compatibility unknown 30 days of GitHub stars
+2 stars
+0% observed change
Oct 14 +0 stars 0 total Oct 15 +0 stars (estimated) 0 total Oct 16 +0 stars (estimated) 0 total Oct 17 +1 stars (estimated) 1 total Oct 18 +0 stars (estimated) 1 total Oct 19 +0 stars (estimated) 1 total Oct 20 +0 stars 1 total Oct 21 +0 stars (estimated) 1 total Oct 22 +0 stars (estimated) 1 total Oct 23 +0 stars (estimated) 1 total Oct 24 +0 stars (estimated) 1 total Oct 25 +0 stars (estimated) 1 total Oct 26 +0 stars (estimated) 1 total Oct 27 +0 stars (estimated) 1 total Oct 28 +0 stars (estimated) 1 total Oct 29 +0 stars (estimated) 1 total Oct 30 +0 stars (estimated) 1 total Oct 31 +0 stars (estimated) 1 total Nov 1 +0 stars (estimated) 1 total Nov 2 +0 stars (estimated) 1 total Nov 3 +1 stars (estimated) 2 total Nov 4 +0 stars (estimated) 2 total Nov 5 +0 stars (estimated) 2 total Nov 6 +0 stars (estimated) 2 total Nov 7 +0 stars (estimated) 2 total Nov 8 +0 stars (estimated) 2 total Nov 9 +0 stars (estimated) 2 total Nov 10 +0 stars 2 total Nov 11 +0 stars 2 total
Oct 14 Oct 28 Nov 12