Add Hindsight long-term project memory to DeepSeek Harness: automatically recall knowledge pages and context each session, conversations auto-saved, shared memory bank per repository.
$ dsh plugin --profile web add @vectorize-io/hindsight-coding-agents
Provides local four-layer long-term memory for DeepSeek Harness (L1 trajectory / L2 strategy / L3 world model / skills), with automatic retrieval each user turn and support for registering six memory tools.
$ dsh plugin --profile web add @memtensor/memos-local-plugin
Provides cloud-based persistent memory for DeepSeek Harness. Automatically retrieves relevant conversation history before each turn and writes to memory after conversations end. Exposes five memory tools that models can invoke.
Equip DSH Web with the Mnemon persistent memory backend, connecting session hot memory and project archives with Mnemon long-term storage as a three-layer supervised memory system.
$ dsh plugin --profile web add github:mnemon-dev/mnemon
Connects DSH to a Git-driven failure experience memory network. When agents encounter errors, they can retrieve 290+ community-verified fix paths, exposed as `deepseek
$ dsh plugin --profile web add github:Ikalus1988/MisakaNet
dsh Shared Core Package: Inserts tools, model adapters, subagent providers, persistence, and policies as a first-layer patch onto the profile root—every profile includes this layer by default.
$ dsh plugin --profile web add @deepseek-ai/dsh-base
Provides auditable long-term memory for DSH Agent based on Noema: cross-session fact storage and retrieval, settings page management, and memory import from 9 other AI tools.
$ dsh plugin --profile web add @zseven-w/dsh-noema
Provides persistent memory and retrieval-augmented context for DeepSeek Harness agent, supporting automatic recall, user input capture, turn summary, and local RAG knowledge base.
Provides a three-layer automatic memory engine for DSH Web GUI: automatic logging of each conversation turn, keyword retrieval, and cross-tool memory inheritance, with visualization panel and calendar view.
$ dsh plugin --profile web add @a9i5k4/dsh-auto-memory
Add a recall button next to DSH user messages to roll back both project files and conversation history to their state before that message was sent, using shadow git repository snapshots with official sessions.fork.
Provides a local project memory layer for DSH Agent: persists work conclusions, lessons learned, and follow-up actions across sessions, with bounded Hot Memory auto-injection + BM25 ranking retrieval + web panel management.
DSH-KRouter — Agent knowledge OS. Self-evolution. First qualifying day auto-provisional; second accepted task → formal. Correction-first. Retrieval is the lock, not the product. Timer off ≠ optional. Cursor, Codex, Claude Code, DeepSeek Harness.
$ dsh plugin --profile web add github:398894496-arch/runtime36
Provides persistent memory for DSH Web GUI using a local Markdown knowledge base, enabling AI to read, write, and search vault files with sidebar directory browsing.
$ dsh plugin --profile web add dsh-client-ui-obsidian-memory
Adds 1-4 pixel kitten skins to the DSH Web interface while preserving the original UI. Kittens are draggable, can change outfits, and respond to interactions.
$ dsh plugin --profile web add --allow-build=dskin github:dancingmemory/dskin
Connect each human-machine turn from DeepSeek Harness to a local TMCRA memory service. Retrieve relevant evidence first, then inject it. Store questions and answers by role, with automatic queue retry on write failure.
$ dsh plugin --profile web add github:reshuibuduo/tmcra-memory#path:integrations/deepseek-harness-local
Self-evolving memory + skill lifecycle for DeepSeek Harness — durable cross-session memory with zero-token deterministic recall, tiered approval, reinforcement learning from repetition, and anti-bloat convergence for both skills and memory.
$ dsh plugin --profile web add github:chenzheshushi-commits/dsh-evolve
Manages hierarchical long-term memory for math notes in Obsidian, including profiles, topics, typed records, and evidence, while supporting theorem indexing and reminders.
Provides task-isolated long-term memory for DeepSeek Harness: each task's memory is stored separately, with memory read, write, retrieval, and prompt injection only effective within the current task.
$ dsh plugin --profile web add github:wangyihao0001-oss/dsh-task-memory
Ported Tencent Cloud's open-source 4-layer memory system (TDAI Agent Memory) into DSH: automatically captures conversations and uses LLM to extract facts, preferences, and profiles; during next conversation, automatically recalls and injects relevant memories based on current message; legacy memory directory remains functional.
Project-scoped cross-session memory: PROJECT.md snapshot injected into the first user message, a memory_remember tool, and auto-reflection after ReAct tasks; each project keeps its own memory file.
$ dsh plugin --profile web add github:Phant0Meow/dsh-memory-meow
Local personality, memory and distillation layer for DeepSeek Harness — SOUL/IDENTITY/USER/STATE persona files, cross-session memory and LLM distillation, auto-loaded on startup.
$ dsh plugin --profile web add github:yuhui-sama/dsh-agentsoul
Lightweight memory plugin for DSH: four Markdown files (USER/PROJECT/WORKLOG/CONVENTION) plus append/distill actions, zero external parts, prefix-cache-friendly dual-layer injection.
Retrieved ≠ injected: CBDC-gated memory for DeepSeek Harness — decides how memory is USED (use/verify/ignore + feedback learning + audit), not just stored
MemoryHub (mh) plugin for DeepSeek Harness (dsh): auto-loads checkpoint memory on session start, adds mh_* tools and the mh skill, and a Memory tab in the web UI
$ dsh plugin --profile web add --allow-build=dsh-memoryhub github:solknight48/dsh-memoryhub
Message recall (撤回) plugin for the DSH Web UI: one undo button on each user message that removes the turn and everything after it, durably, without reverting code changes.
Generic stateless dialogue compression that mimics human memory. LLM conversations stay bounded forever — no session management, no context overflow. Ships a ready-to-use adapter for DeepSeek Harness (DSH).
$ dsh plugin --profile web add github:TuringCorp-net/mosaic_compress
Unified fleet-wide agent memory system for DeepSeek Harness — shared Obsidian Vault for dsh, Codex, Claude Code & Hermes with dependency-free Python FTS5 core.
$ dsh plugin --profile web add dsh-unified-agent-memory
Brain-inspired cognitive plugin: episodic, semantic, procedural and emotional memory with reinforcement-learning signals, a goal stack with time triggers, curiosity-driven autonomous web research, and proactive initiatives.
Cross-session persistent cost ledger for DeepSeek Harness: logs every LLM token usage to SQLite and exposes record/query/budget tools. Built-in DeepSeek pricing, overridable via config.
DSH Cordis plugin: turn any source text into an AI knowledge graph (facts/inferences/concepts/definitions/examples/counter-examples/rules) with two-way linking between the graph and the original text.
$ dsh plugin --profile web add github:cwbcheng/dsh-knowledge-graph