Project-scoped persistent key-value memories with set/get/list/delete/search model tools and JSON-file storage.
Install
# from GitHub (first run asks for allowBuilds approval — follow the hint, retry)
dsh plugin --profile web add github:DamonKoy/dsh-plugins#path:/packages/dsh-memories
Any plugin you install runs third-party code with your own permissions — it can read your files, use your credentials, and reach the network, and tool approvals don’t sandbox it. GitHub-sourced plugins also run build scripts at install time — pnpm blocks those until you allow them, so an install can stop with ERR_PNPM_GIT_DEP_PREPARE_NOT_ALLOWED or ERR_PNPM_IGNORED_BUILDS; dsh prints the exact key to add under allowBuilds in your profile’s pnpm-workspace.yaml, and the install works on the next run. Allowing a build is a trust decision: only install sources you trust, and pin a commit (github:owner/repo#sha).
README
Project-scoped memory system for DeepSeek Harness. Inspired by Codex 0.145/0.146 memories (persisted names, project-scoped memories, efficient resume).
English | 中文
What it does
- Persistent key-value memories stored per scope in
~/.dsh/dsh-memories/<scope>.json— survive across sessions and restarts. - Model tools:
memory_set(scope, key, value)— upsert one memorymemory_get(scope, key)— read one memorymemory_list(scope)— list memories (newest first)memory_delete(scope, key)— remove one memorymemory_search(scope, query)— substring search over keys and values
- Client RPC:
memories/scopes,memories/list,memories/get,memories/upsert,memories/remove— ready for a settings/memory panel UI.
Use scopes = project names (e.g. my-project) to keep per-project context
separate; omit scope for global memories.
Install
dsh plugin --profile web add github:DamonKoy/dsh-plugins#path:/packages/dsh-memories
Restart dsh web.
Storage
Plain JSON files, one per scope:
{
"entries": {
"deploy-command": {
"value": "pnpm run deploy:prod",
"createdAt": 1786000000000,
"updatedAt": 1786000000000
}
}
}
Scopes are sanitized to [A-Za-z0-9._-] and capped at 80 chars; global is
the fallback scope. Corrupted files fall back to an empty store instead of
crashing the plugin.
Security notes
- Memories are plaintext local files in your home directory (mode honors
your umask). Do not store credentials — use the credential service or
dsh-secret-redactorfor sensitive material. - No data ever leaves the machine.
Roadmap
- Prompt-section injection of the current scope's summary at session start.
- Memory panel UI (list / edit / delete) via the client RPC.
- Automatic scope from the session workspace instead of explicit arguments.
License
MIT
Links
More in this category
vectorize-io/hindsight#coding-agents★ 44359
Hindsight, agent memory that learns: long-term project memory with auto recall and retain, knowledge pages, deep reflection, and per-repo memory banks.
volcengine/OpenViking#examples/dsh-memory-plugin★ 39101
OpenViking memory and context bundle for DeepSeek Harness: pre-step auto-recall and profile injection, session capture, `viking://` URI guarding, and recall/write memory tools backed by an OpenViking server.
agentscope-ai/ReMe#dsh★ 3545
Connects DeepSeek Harness to ReMe's local-first, self-evolving personal knowledge base: automatically captures completed main-agent conversations as user-owned Markdown memory, searches conversations and source material through reme_search with BM25, optional embeddings, and wikilink expansion, and schedules daily memory consolidation.
zilliztech/memsearch#MemSearch★ 2691
Shared Markdown memory for DSH and other coding agents, with automatic capture, pre-step context injection, searchable recall, and memory-to-skill self-evolution through a review panel.
vshulcz/deja-vu#extensions/dsh★ 1116
Reads the session files thirty-three other coding agents on this machine already wrote — Claude Code, Codex, Cursor, VS Code Copilot Chat, opencode, OpenClaw, Hermes, Kimi, Cline, Zed and more — including sessions from before it was installed: six tools (deja_recall, deja_session, deja_blame, deja_fix, deja_how, deja_remember), a /deja command, and optional automatic recall added to the runtime context. Local BM25 index, no LLM, no embeddings, no network (dsh plugin --profile web add dsh-deja).
adoresever/graph-memory★ 633
Traceable, searchable cross-session memory for DeepSeek Harness — conversation knowledge as typed graph nodes (TASK/SKILL/EVENT) and typed edges.
Community comments
Comments are public GitHub Discussions. Loading them connects to GitHub and Giscus; a GitHub account is required to post.