dsh-snippets
Run the following command in DeepSeek Harness:
dsh plugin install taxueseek/dsh-snippets
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install taxueseek/dsh-snippets in DeepSeek Harness to install it; the source code is available at https://github.com/taxueseek/dsh-snippets
About this plugin
You paste the same docker logs command every day, reuse the same prompt in every AI chat, and repeat the same regex pattern in every terminal session. These high-frequency snippets should be one call away, yet they scatter across chat logs, sticky notes, and your own memory. dsh-snippets solves the classic pain of saving something once and never finding it again — a single tool call in any session brings it straight back.
The design philosophy is five verbs and one file. snippet_save, snippet_search, snippet_get, snippet_delete, snippet_count: the semantics are obvious at a glance, with no extra concepts to learn. All data lives in a single JSONL file at ~/.dsh/snippets/snippets.jsonl, one JSON object per line, human-readable, manually editable, and trivially backed up. No database, no vector store, no web panel; the only runtime dependency is the harness SDK. Even more important is the explicit-memory principle: the AI will not silently archive your conversations into its own memory store. A snippet is written only when you explicitly ask it to be, and you always know exactly what was saved, where it lives, and what it looks like.
It is built for terminal-heavy users who refuse to spin up another background service, for minimalists who believe one well-structured file beats an entire tech stack, and for anyone who values data sovereignty and does not want an AI quietly indexing their chat history. If your personal collection of go-to commands, code blocks, and prompts runs into the dozens rather than the thousands, dsh-snippets is exactly the right amount of tooling — no more, no less.
Use Cases
- Retrieve a previously saved command or code block with a single tool call in any session
- Reuse prompts and regex snippets across sessions without re-typing or re-searching
- Manually edit, review, and back up a personal snippet collection in a local JSONL file
Best For
- Terminal-heavy users who do not want another database or background service
- Minimalists who prefer managing all snippets in one human-readable file with full transparency
- Developers who want explicit control over what the AI stores and retrieves, keeping only actively curated content
Related Plugins
Traceable, searchable cross-session memory for AI agents that turns conversation knowledge into a typed knowledge graph and recalls relevant subgraphs instead of replaying full history, natively integrated with DeepSeek Harness.
Gives DSH AI cross-session long-term memory, to-do and skill management, plus multi-session orchestration, external AI delegation, and an infinite canvas that grows with you.
dsh-mnemon is a three-tier, pluggable, Agent-driven memory system for DeepSeek Harness, combining Runtime memory, Project Documents, and replaceable Memory Spaces with nine long-term providers.
An opinionated, zero-infrastructure file-based memory protocol: human-readable markdown files plus a bounded always-loaded index, curation discipline, and no database or embeddings.