nowledge-mem

ferstar's avatarfrom ferstar

Unified CLI for personal knowledge base. This skill should be used when users want to save memories, search past conversations, persist current sessions, or manage stored information. Triggers: 记忆, 知识库, 保存会话, 搜索记忆, recall, remember, save memory, search knowledge, persist session, nm命令.

⭐0stars🔀0forks📁View on GitHub🕐Updated Dec 27, 2025

When & Why to Use This Skill

Nowledge Mem is a powerful CLI-based personal knowledge base and memory management skill for Claude. It enables users to seamlessly capture, search, and persist conversation histories and critical information, ensuring long-term context retention and efficient knowledge retrieval across different sessions.

Use Cases

  • Long-term Context Retention: Save specific technical insights, project decisions, or personal preferences during a chat to ensure they are available for future interactions.
  • Efficient Knowledge Retrieval: Use the 'nm search' command to instantly find past solutions, code snippets, or specific data points stored in your personal knowledge base.
  • Session Persistence: Persist the state of complex, multi-step project sessions to maintain continuity and prevent the loss of progress in long-running workflows.
  • Structured Information Management: Organize stored memories using titles, importance ratings, and custom labels to build a highly structured and searchable digital brain.
namenowledge-mem
description|
Unified CLI for personal knowledge base. This skill should be used when users want to save memories, search past conversations, persist current sessions, or manage stored information. Triggers记忆, 知识库, 保存会话, 搜索记忆, recall, remember, save memory, search knowledge, persist session, nm命令.

Nowledge Mem

Unified CLI for Nowledge Mem knowledge base management.

Quick Reference

Command Description
nm search "..." Search memories
nm add "..." Add a new memory
nm persist Save current session
nm expand <id> View full thread
nm update <id> Update memory
nm delete <id> Delete memory
nm labels List all labels
nm diagnose Check connectivity

Execution

Run all commands via uv run nm from the skill directory.

cd <skill-directory>
uv run nm <command> [options]

Core Workflows

Search Knowledge Base

# Basic search
uv run nm search "Python async patterns"

# Verbose with JSON output
uv run nm search "API design" --verbose --json

Add Memory

# Simple add
uv run nm add "Content to remember"

# With metadata
uv run nm add "Important config" --title "DB Config" --importance 0.8 --labels "config,db"

Persist Current Session

Set PROJECT_PATH to the actual project directory before running.

PROJECT_PATH=/path/to/project uv run nm persist
PROJECT_PATH=/path/to/project uv run nm persist --title "Feature implementation"

Expand Thread

uv run nm expand <thread_id>

Reference Documentation

For detailed parameters and options:

  • Command parameters: references/command_reference.md
  • Configuration: references/configuration.md
  • Usage patterns: references/usage_patterns.md
  • Troubleshooting: references/troubleshooting.md

Trigger Keywords

Language Keywords
中文 记忆, 知识库, 保存, 搜索, 记录, 存储, 会话, 持久化
English memory, knowledge, save, search, persist, recall, remember, session