Can I distinguish between good and bad patterns?

Yes, you can use the --success or --failed flags to explicitly categorize entries as proven best practices or avoided anti-patterns.

Does it require a cloud subscription or database?

No, it requires only the memory MCP server and works entirely locally with zero configuration required to start saving patterns.

Is this memory shared across different AI agents?

By default, memory is scoped, but you can use the --agent flag to specify a specific agent context or the --global flag to share memory across multiple projects.

How does the remember skill store information?

It uses a local knowledge graph via the MCP memory server, allowing for zero-config, persistent storage of entities and relationships directly on your machine.

How does entity extraction work?

The skill automatically identifies capitalized technologies, hyphenated agent names, and relationship patterns like 'X uses Y' to create a structured map of your stack.

Remember Knowledge Graph

Name: Remember Knowledge Graph
Author: yonatangross

byyonatangross

•

Productivity & Workflow

Persists architectural decisions, coding patterns, and lessons learned into a local knowledge graph for persistent project memory.

The remember skill enables Claude to build a structured Best Practice Library by capturing critical project context, success/failure outcomes, and technical relationships. By automatically extracting entities and mapping dependencies within a zero-config knowledge graph, it ensures that previous decisions—such as choosing specific libraries or identifying anti-patterns—are preserved across sessions to prevent repetitive mistakes and accelerate development workflows.

Key Features

01Context-aware categorization for decisions, architecture, and performance patterns

0292 GitHub stars

03Duplicate detection and intelligent conflict warnings for existing graph nodes

04Automated entity and relationship extraction for structured knowledge graph storage

05Tracking of success and failure outcomes to build a verified Best Practice Library

06Agent-scoped and global memory options for cross-project consistency

Use Cases

01Documenting anti-patterns that caused bugs or performance issues to prevent future recurrence

02Recording why a specific library or architecture was chosen over another to preserve design intent

03Mapping technology dependencies and agent-specific preferences for team-wide alignment

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add yonatangross/orchestkit remember

For use in Claude.ai and ChatGPT

Download Skill