Can I filter the memory search results?

Yes, the tool supports optional label filtering, such as 'backend' or 'architecture', to narrow down the search to specific domains.

How does the skill decide when to search?

The skill uses autonomous recognition to trigger searches when it detects continuity with prior work, pattern matches with past issues, or implicit recall signals like 'like we did before'.

What is the difference between 'normal' and 'deep' modes?

'Normal' mode is the default for fast retrieval, while 'Deep' mode provides a more comprehensive search when complex context or thorough recall is required.

How do I install the necessary MCP for this skill?

Use the command: 'claude mcp add --transport http nowledge-mem http://localhost:14242/mcp --scope user' to add the required Model Context Protocol.

What is the primary function of the Search Memory skill?

It allows Claude Code to proactively search your stored historical insights for relevant past decisions and solutions to improve current task accuracy.

Search Memory

Name: Search Memory
Author: nowledge-co

bynowledge-co

•

生产力与工作流

Retrieves relevant past insights, architectural decisions, and solved patterns from a memory store to provide contextually aware AI responses.

The Search Memory skill empowers Claude Code with a long-term contextual layer, allowing it to autonomously recognize when past solutions or technical decisions are relevant to your current task. By proactively searching stored knowledge during debugging, architecture discussions, or recurring workflows, it ensures continuity across sessions and prevents the repetition of past mistakes. It utilizes semantic search with adjustable depth and relevancy scoring, enabling Claude to recall why specific choices were made and apply those breakthroughs to new challenges without requiring manual prompting.

主要功能

0112 GitHub stars

02Multi-mode Semantic Search (Normal and Deep discovery)

03Autonomous Search Recognition based on conversational context

04Relevancy Scoring to filter high-confidence matches

05Multi-language aware terminology preservation

06Implicit Recall capability for vague references like 'that approach from before'

使用场景

01Retrieving root cause analyses for recurring bugs and complex debugging patterns

02Recalling specific architectural rationale and decisions made in previous sessions

03Maintaining consistency in coding standards and domain-specific conventions across projects

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add nowledge-co/community search-memory

For use in Claude.ai and ChatGPT

Download Skill