dsh-memory
Lingshu long-term auditable memory system and cognitive graph for DeepSeek Harness.
Install
$ dsh plugin --profile web add .Plugin Overview & Capabilities
AI-assisted organization based on the public repository snapshot. The content must be grounded in source evidence and does not replace compatibility or security verification.
Key Capabilities
- Cross-session long-term memory storage backed by plain Markdown cognitive graphs
- Deterministic rule-based write/recall gatekeeping with zero blackbox LLM hallucinations
- CCG six-element structured memory compilation with external review verification
- Multi-harness MCP support covering DeepSeek Harness, Claude Code, Codex CLI, and others
- High-performance retrieval engine with Chinese semantic normalization and query bridging
- Role-based access control, write token authentication, and full-trail audit logging
Useful For
- Retaining and recalling user preferences, project conventions, and decisions across DSH sessions
- Building auditable and verifiable knowledge graphs for multi-agent collaboration
- Enforcing deterministic memory curation and preventing conversational hallucination in agent workflows
Who It Fits
- DeepSeek Harness developers building stateful autonomous agents
- Engineers requiring strictly auditable and reproducible memory architectures
- Multi-agent platform developers needing shared MCP memory infrastructure
Documented Limitations
- Requires Node.js >= 22.19 and a local Python environment to run the md_cg brain backend
- Default execution is read-only guest mode unless configured with an issued write token (MDCG_TOKEN)
- Plain text memory writes must satisfy CCG six-element formatting or strict policy rules to avoid gate rejection
DSH Compatibility
Version-specific runtime evidence collected by DSH Plugin. A missing result means we have not tested that combination yet.
Security Signals
Objective signals discovered from package metadata and source inspection. These are not a guarantee that a plugin is safe.
package.json declares 0 runtime, 9 development, 6 peer, and 0 optional dependencies.
package.json declares DSH bundle metadata.
GitHub reports the repository license as MIT.
A root package.json was captured and can be inspected by the registry.
package.json declares a prepare lifecycle script that may run during relevant package installation workflows.
Public GitHub source metadata is available for this registry snapshot.
Source & Registry Notes
Public provenance, Registry classification, and the latest source check for this entry, kept separate from runtime verification.
- Source repository
- FuRongJun-1999/dsh-memory
- Registry source
- GitHub · dsh-plugin topic
- Registry classification
- Plugin
- Source checked
- dedef08 · 2026-10-06
This project is independently indexed from public source information. DSH Plugin is not affiliated with DeepSeek or the plugin author. Always check the author repository before installation.
Repository Activity
- GitHub stars
- 324GitHub stars
- Forks
- 24
- Open issues
- 3
- Last commit
- 2026-10-06
- Last release
- 2026-10-06
Related DSH Plugins
Ranked by overlapping capabilities, use cases, plugin type, categories, and DSH profile.
Composable, view-based layered memory plugin bundle for DeepSeek Harness with pluggable sources and providers.
View plugin →LLM-supervised persistent memory and graph-based recall integration bundle for DeepSeek Harness.
View plugin →Shared local-first memory and session history indexer for DeepSeek Harness and 30+ coding agents.
View plugin →Second-model AI auto-review for DeepSeek Harness approval requests with fail-closed verdicts and audit logging.
View plugin →