DeepSeek Harness Plugins
DeepSeek Harness PluginManifest Valid

graph-memory

Traceable, searchable, cross-session knowledge graph memory engine for DeepSeek Harness.

VisionUI & ProductivityDeveloper ToolsBrowser & WebSecurity & PolicySkills & WorkflowsRemote Execution
584GitHub Stars85ForksUpdated2026-08-31

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.

source-grounded
Graph Memory is a local-first cross-session memory and context engine native to DeepSeek Harness. It structures conversation history into typed knowledge graphs (TASK, SKILL, EVENT nodes connected by semantic edges), replacing indiscriminate prompt history replay with targeted subgraph retrieval via vector search, FTS5 lexical matching, community detection, and Personalized PageRank.

Key Capabilities

  • Durable cross-session memory storage using local SQLite
  • Dual-path recall combining semantic vector search and FTS5 lexical fallback
  • Graph ranking via community detection and Personalized PageRank (PPR)
  • DSH CompactionEngine integration preserving source message provenance while bounding prompt context
  • Deterministic manual memory recording and search via gm_record, gm_search, and gm_status tools
  • Configurable message retention policies with dry-run support

Useful For

  • Retaining developer workflows, past solutions, and debugging steps across multiple DSH sessions
  • Compressing long conversation contexts while preserving traceable provenance and graph relationships
  • Recalling agent skills and task dependencies without manual copy-pasting between chats

Who It Fits

  • DeepSeek Harness power users requiring persistent cross-session recall
  • AI agent developers seeking structured long-term memory architectures
  • Teams needing local-first, privacy-compliant agent knowledge storage

Documented Limitations

  • Automatic knowledge extraction depends on the stability and quality of the LLM output
  • Pro visual workbench currently offers a read-only card snapshot client without 2D/3D or drag-and-drop context injection
  • Requires optional embedding endpoint configuration for semantic vector search; falls back to FTS5 otherwise

DSH Compatibility

Version-specific runtime evidence collected by DSH Plugin. A missing result means we have not tested that combination yet.

Not tested yetNo runtime compatibility tests have been published yet.

Security Signals

Objective signals discovered from package metadata and source inspection. These are not a guarantee that a plugin is safe.

Dependency Counts

package.json declares 1 runtime, 5 development, 3 peer, and 0 optional dependencies.

info
Dsh Bundle Declared

package.json declares DSH bundle metadata.

info
License Declared

GitHub reports the repository license as MIT.

info
Package Manifest Available

A root package.json was captured and can be inspected by the registry.

info
Source Available

Public GitHub source metadata is available for this registry snapshot.

info

Source & Registry Notes

Public provenance, Registry classification, and the latest source check for this entry, kept separate from runtime verification.

Source repository
adoresever/graph-memory
Registry source
GitHub · dsh-plugin topic
Registry classification
Plugin
Source checked
bc5c0dd · 2026-08-31

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
584GitHub stars
Forks
85
Open issues
12
Last commit
2026-08-31
Last release
2026-03-18
For maintainers

Maintaining this plugin?

This listing is generated from public repository data. If you maintain this project, you can review the information and share this listing with your users if you find it useful.

Add to README
Listed on DSHPlugin.app