dsh-memory

Durable cross-session memory for DeepSeek Harness: local SQLite FTS5 store, three model-facing tools, and a recall prompt section — no embedding service and no sidecar

AI Analysis

核心用途是为 AI 提供轻量级的跨会话长期记忆。适合需要 Agent 记住历史设定、项目背景或用户偏好的长期任务，无需复杂的向量数据库配置。

Package

dsh-memory

Version

0.1.0

License

MIT

Last updated

Aug 13, 2026

Configuration

- id: memory
  name: dsh-memory
  config:
    path: !!js dshHomePath('memory/memory.db')
    promptRecentCount: 10
    promptMaxChars: 2000
    maxTextChars: 2000
    searchLimitDefault: 10
    searchLimitMax: 50
    promptOrder: 50

Field	Default	Meaning
`path`	— (required)	SQLite file, or `:memory:` for an ephemeral store
`promptRecentCount`	`10`	Unpinned recent memories offered to the prompt section
`promptMaxChars`	`2000`	Budget for the rendered section; overflow is reported as a count, and pinned memories are emitted first so they survive a tight budget
`maxTextChars`	`2000`	Maximum characters accepted for one memory
`searchLimitDefault`	`10`	`memory_search` limit when the model omits it
`searchLimitMax`	`50`	Hard cap, whatever the model asks for
`promptOrder`	`50`	Section order; `-100` is the harness identity, `0` the persona

path has no code-side default on purpose: a default would scatter durable user facts into whatever directory the harness happened to start in. The deployment value lives in the patch row.

dsh-memory

AI Analysis

Install

README

Configuration