Delete topic learning files and related knowledge graph entries.
npx skills add https://github.com/ourines/claude-code-learn --skill forget
Please help me install this Agent Skill. Skill: forget Source: https://github.com/ourines/claude-code-learn/tree/main/skills/forget Command: npx skills add https://github.com/ourines/claude-code-learn --skill forget
Deleting outdated or unwanted topic learnings from the local memory store and, if possible, removing corresponding graph entries to keep Claude's knowledge base tidy and accurate.
Delete the knowledge file for a topic from ~/.claude/learnings/ and remove its graph entry if available.
High-intent search queries and answers about installing and using this skill.
To delete saved Claude learnings from local storage, remove the specific topic's learning file from the ~/.claude/learnings/ directory and optionally remove its corresponding entity from the knowledge graph using memory tooling if available.
The best way to clean up the knowledge graph when removing topic-specific data is to use memory tooling like mcp__memory__search_nodes and mcp__memory__delete_entities to find and purge corresponding graph entries alongside local file deletion.
Yes, you can list all saved learnings to preview them before deletion, allowing you to verify which topic-specific files in the local memory store you intend to purge safely.
Knowledge graph cleanup requires memory tooling like mcp__memory__delete_entities to remove graph entries; without it, the process performs a safe local file deletion only and skips graph updates.
You need to purge outdated topic learnings when your local memory store and knowledge graph contain inaccurate or unwanted data, requiring cleanup to keep Claude's knowledge base tidy and accurate.
rhpdsConfigure persistent Memory MCP knowledge graph storage for Claude Code sessions.
an8079Build and query Neo4j knowledge graphs with SPARQL for RAG systems.
unclehowellOrganize knowledge into PARA-based memory files with three-layer structure.
nathanvaleGuides building a knowledge management system using the PARA method and CODE workflow.
