continuous-learning

Automatically extract reusable patterns from Claude Code sessions and save them as learned skills for future use.

2.27x

Quality

47%

Does it follow best practices?

Impact

100%

2.27x

Average score across 3 eval scenarios

Securityby

Risky

Do not use without reviewing

Optimize this skill with Tessl

npx tessl skill review --optimize ./skills/continuous-learning/SKILL.md

Evaluation results

100%

10%

Automate Knowledge Capture from Claude Code Sessions

Hook configuration for automatic session learning

Criteria

Without context

With context

Stop hook event

100%

Command hook type

100%

evaluate-session.sh command

100%

continuous-learning directory

100%

Wildcard matcher

100%

Correct nesting structure

100%

Valid JSON output

100%

58%

Build a Session Analysis Trigger Script

Session evaluator script implementation

Criteria

Without context

With context

Transcript path from stdin JSON

100%

CLAUDE_TRANSCRIPT_PATH fallback

100%

Silent exit on missing transcript

50%

100%

Counts user messages only

100%

Minimum session length of 10

100%

Skips short sessions

70%

100%

Uses jq for config parsing

100%

Config loaded from SCRIPT_DIR

100%

Learned skills path

100%

Creates output directory

100%

Configure Pattern Extraction for Session Learning

Pattern extraction configuration

Criteria

Without context

With context

error_resolution pattern

100%

user_corrections pattern

100%

workarounds pattern

100%

debugging_techniques pattern

100%

project_specific pattern

100%

simple_typos ignored

100%

one_time_fixes ignored

100%

external_api_issues ignored

100%

auto_approve false

100%

extraction_threshold medium

100%

Repository: affaan-m/everything-claude-code
Commit: 5df943e

Evaluated: about 1 month ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Automate Knowledge Capture from Claude Code Sessions Build a Session Analysis Trigger Script Configure Pattern Extraction for Session Learning

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.