Search Memory

nowledge-co/community/nowledge-mem-droid-plugin/skills/search-memory

作者 nowledge-co1f4d7eb61466dbfc93ce8d7a0a806c4d4a196dd0無授權條款收錄於 2026年10月9日更新於 2026年10月9日

Search memory and thread history when past knowledge would materially improve the answer. Route between distilled memories and prior discussions instead of treating them as separate silos.

僅含說明AI & Agents
AI 產生的概覽

引導代理在過往脈絡有助於回答時,搜尋已儲存的記憶與先前的討論紀錄。

功能
此技能提供判斷何時應檢索持久記憶或早期對話串的說明。它列出強訊號與情境訊號,提供記憶搜尋、對話串搜尋以及逐步取得對話串內容的指令,並訂定揭露所回想知識的回應約定。它產出的是檢索指引與行為規範,而非檔案或程式碼。
適用情境
當任務與先前工作相關、與之前解決過的問題相似,涉及審查、回歸、發布、文件對齊或連接器行為,或使用者詢問某項決策的原因時使用。也適用於「那個做法」「像之前那樣」等隱含回想的說法。對於全新任務、一般語法問題或使用者希望獲得全新觀點的情況則應略過。
執行需求
需要 nmem 命令列工具及其記憶與對話串搜尋子命令,並能存取底層的記憶與對話串儲存。此技能未附帶指令碼,僅為說明性指示。

Search Memory

When to Search

Strong signals — search when:

  • the current task connects to prior work
  • the bug or design resembles something solved earlier
  • the task is a review, regression, release, docs-alignment, or connector-behavior question
  • the user asks why a decision was made
  • a previous discussion or session likely contains the missing context
  • the user uses implicit recall language: "that approach", "like before"

Contextual signals — consider searching when:

  • complex debugging where prior context would narrow the search space
  • architecture discussion that may intersect with past decisions
  • domain-specific conventions the user has established before

Skip when:

  • the task is fundamentally new
  • the question is generic syntax or reference material
  • the user explicitly wants a fresh perspective

Tool Usage

Start with durable knowledge:

bash
nmem --json m search "3-7 core concepts"

Use thread search when the user is really asking about a prior conversation:

bash
nmem --json t search "query" --limit 5

If a memory result includes source_thread, or thread search identifies the likely discussion, fetch progressively:

bash
nmem --json t show <thread_id> --limit 8 --offset 0 --content-limit 1200

Increase --offset only when more messages are actually needed.

Response Contract

  • Be explicit when you relied on recalled knowledge
  • Prefer the smallest retrieval surface that answers the question
  • Suggest distillation only if the current discussion produced new durable knowledge
  • For continuation-heavy engineering work, search near the start of the task rather than waiting for an explicit recall request

來源與署名

來源:nowledge-co/community位於nowledge-mem-droid-plugin/skills/search-memory提交1f4d7eb

授權條款: 無授權條款

內容歸原作者所有。SourceWeft 從公開儲存庫中收錄這些內容。

檢舉或申請下架