continuous-learning

Extract and persist learning patterns from Claude Code sessions.

17|1|Updated Dec 7, 2025
One-click install
npx skills add https://github.com/mhylle/claude-skills-collection --skill continuous-learning-mhylle
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: continuous-learning
Source: https://github.com/mhylle/claude-skills-collection/tree/main/skills/continuous-learning
Command: npx skills add https://github.com/mhylle/claude-skills-collection --skill continuous-learning-mhylle

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a systematic framework to extract, organize, and persist valuable patterns discovered during work sessions, ensuring insights don't vanish between sessions.

Core Features & Use Cases

  • Pattern detection and extraction from session data (errors, corrections, workarounds, debugging techniques, project-specific patterns)
  • Persistent storage under ~/.claude/skills/learned/ with indexing for quick retrieval
  • Automatic triggers on Stop hook or explicit user invocation to surface reusable knowledge

Quick Start

To start capturing learnings from your current session, say "save learnings from this session" and check the summary afterward.

Frequently Asked Questions about continuous-learning

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I save learnings from a Claude Code session so they persist?

To save learnings from a Claude Code session, you can explicitly invoke the extraction process or rely on automatic triggers like Stop hooks and timeouts to capture and persist patterns for reuse.

What types of session patterns can be extracted for knowledge management?

Extractable session patterns include error resolutions, user corrections, workarounds, debugging techniques, and project-specific patterns, all organized by pattern type for future knowledge management.

How does deduplication work when storing extracted learning patterns?

Deduplication prevents redundant entries when storing extracted learning patterns by comparing new session data against existing records, ensuring the persisted knowledge base remains concise and efficient.

Can I configure the extraction threshold for capturing session patterns?

Yes, you can configure the extraction threshold for capturing session patterns by modifying the config.json file, allowing you to control the sensitivity of pattern detection during your work.

Where are extracted learning patterns stored for later retrieval?

Extracted learning patterns are stored under the ~/.claude/skills/learned/ directory with indexing, providing a persistent storage structure that ensures quick retrieval of reusable knowledge across sessions.