Is my data shared with other users?

No. All observations and instincts are stored locally on your machine. You have full control over what patterns are kept and which are exported for sharing.

What is an 'atomic instinct'?

An atomic instinct is a single, focused behavior pattern—such as a preference for a specific library—assigned a confidence score based on how often it's observed.

How do instincts turn into skills?

The system uses an '/evolve' command that analyzes your collection of personal instincts and clusters related patterns into more complex, permanent skills or agent roles.

How does Continuous Learning v2 differ from standard Claude skills?

Unlike standard skills which are triggered probabilistically, v2 uses deterministic hooks to ensure 100% of tool interactions are observed and learned from systematically.

Continuous Learning v2

Name: Continuous Learning v2
Author: Nixdorfer

byNixdorfer

•

Productividad y Flujo de Trabajo

Transforms developer interactions into reusable AI instincts that evolve into persistent skills and specialized agents.

Continuous Learning v2 is an advanced instinct-based architecture designed to turn Claude Code sessions into a persistent knowledge base. By utilizing deterministic hooks to monitor tool usage and user corrections, the system identifies 'atomic instincts'—small, weighted behaviors that reflect a developer's specific style, preferences, and error resolutions. Over time, these instincts cluster and evolve into higher-level skills, commands, and specialized agents, ensuring Claude becomes increasingly tailored to your unique workflow without the need for manual prompt engineering.

Características Principales

01Knowledge Portability: Support for importing and exporting learned instincts across different project environments.

02Automated Evolution: Clusters related instincts to generate complex commands and specialist agents.

03Deterministic Hook Observation: Uses PreToolUse and PostToolUse hooks to capture 100% of interactions.

04Background Pattern Detection: Analyzes workflows using lightweight models to identify recurring patterns without lag.

051 GitHub stars

06Atomic Instinct Modeling: Tracks granular behaviors with confidence scoring from 0.3 to 0.9.

Casos de Uso

01Building a self-improving repository of common error resolutions to prevent repeated troubleshooting.

02Scaling expert developer workflows into reusable commands for improved team productivity.

03Codifying project-specific style guides and architectural preferences automatically through session observation.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add nixdorfer/claudecodetool continuous-learning-v2

For use in Claude.ai and ChatGPT

Download Skill