continuous-learning-v2

Instinct-based learning system that observes sessions via hooks, creates atomic instincts with confidence scoring, and evolves them into skills/commands/agents.

1.58x

Quality

26%

Does it follow best practices?

Impact

89%

1.58x

Average score across 6 eval scenarios

Securityby

Advisory

Suggest reviewing before use

Optimize this skill with Tessl

npx tessl skill review --optimize ./docs/zh-TW/skills/continuous-learning-v2/SKILL.md

Evaluation results

100%

88%

Onboarding Setup for Continuous Learning System

System setup and configuration

Criteria

Without context

With context

PreToolUse hook present

50%

100%

PostToolUse hook present

50%

100%

No Stop hook used

100%

Correct observe.sh path

100%

homunculus base dirs

100%

evolved subdirs

100%

observations.jsonl created

100%

Config version field

100%

Observation size limit

100%

Archive retention

100%

Confidence thresholds

100%

Confidence decay rate

100%

Observer model

100%

Observer interval

100%

Evolution cluster threshold

100%

21%

Capturing Developer Patterns as Instinct Files

Instinct file format and creation

Criteria

Without context

With context

YAML front matter present

50%

100%

id field format

100%

trigger field present

100%

confidence in range

75%

100%

Confidence calibrated to evidence

100%

domain field present

100%

source field present

100%

Atomic: one action per instinct

100%

Evidence section present

100%

Action section present

100%

Correct output location

100%

Five instincts created

100%

68%

-11%

Evolving Accumulated Instincts into Reusable Skills

Instinct evolution and clustering

Criteria

Without context

With context

Cluster threshold enforced

Testing cluster evolved

100%

Git cluster evolved

50%

Correct evolved/ placement

100%

Subdirectory type matches cluster

100%

evolution-plan.md present

100%

Instinct ids referenced

100%

Evolved file is consolidated

100%

Confidence decay noted

80%

70%

No merged instinct files removed

100%

Repository: ysyecust/everything-claude-code
Commit: 79cc4e3

Evaluated: 27 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Onboarding Setup for Continuous Learning System Capturing Developer Patterns as Instinct Files Evolving Accumulated Instincts into Reusable Skills

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.