Does the system handle sensitive information?

Yes, the skill includes specialized commands to scan for sensitive data and supports pre-commit sanitization hooks to protect your privacy.

How does automated recall work?

The system uses a PreToolUse hook to detect when you are facing a problem and automatically suggests relevant lessons or decisions from your graph before you start solving.

What happens if I try to capture a duplicate lesson?

The skill performs a search-before-capture strategy, offering to merge new insights into existing entries or create 'Related' links to maintain a clean knowledge structure.

Can I use this knowledge graph across different LLM platforms?

Yes, the system is platform-agnostic and includes a sync agent to align configuration and knowledge between Claude Desktop, Claude Code, and other assistants.

How does this differ from standard project documentation?

Unlike static docs, this skill automates capture based on real-time interactions and provides context-aware recall directly within your LLM workflow, ensuring knowledge is used when needed.

Knowledge Graph Manager

Name: Knowledge Graph Manager
Author: technomensch

bytechnomensch

•

学習とドキュメンテーション

Builds and manages a platform-agnostic knowledge base to capture lessons, decisions, and institutional memory across development sessions.

This skill provides a structured framework for managing institutional memory within Claude Code by coordinating context, logic, and lifecycle layers to automate knowledge capture and retrieval. It allows developers to persist lessons learned from debugging, architectural decisions, and process improvements into a searchable graph, ensuring that valuable insights are not lost between sessions. By using automated hooks and specialized agents, the skill proactively identifies capture opportunities and offers intelligent recall, preventing the redundant effort of re-solving previously encountered problems.

主な機能

01Proactive duplicate detection to maintain a single source of truth

02Context-aware search and retrieval via a dedicated recall-agent

03Four-layer architecture for seamless data persistence and cross-platform sync

044 GitHub stars

05Automated knowledge capture triggered by specific development patterns and keywords

06Standardized templates for lessons, architectural decisions, and session summaries

ユースケース

01Generating automated session summaries to maintain a clear audit trail of development progress

02Documenting non-obvious bug fixes and root cause analyses to prevent future debugging overhead

03Recording Architectural Decision Records (ADRs) to track technical trade-offs and design choices

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add technomensch/knowledge-graph knowledge-graph-usage

For use in Claude.ai and ChatGPT

主な機能

01Proactive duplicate detection to maintain a single source of truth

02Context-aware search and retrieval via a dedicated recall-agent

03Four-layer architecture for seamless data persistence and cross-platform sync

044 GitHub stars

05Automated knowledge capture triggered by specific development patterns and keywords

06Standardized templates for lessons, architectural decisions, and session summaries

ユースケース

01Generating automated session summaries to maintain a clear audit trail of development progress

02Documenting non-obvious bug fixes and root cause analyses to prevent future debugging overhead

03Recording Architectural Decision Records (ADRs) to track technical trade-offs and design choices