What security features does CORE provide?

CORE enforces a strict separation between private and public repositories, includes built-in prompt injection defenses, and mandates a multi-step check before pushing code to ensure sensitive data is never exposed.

What is the PAI CORE skill?

CORE is the foundational skill for the Personal AI Infrastructure (PAI). It defines the AI's identity, mandatory response patterns, and core operating principles to ensure consistent behavior across all Claude Code sessions.

How does the mandatory response format work?

CORE requires every task-based response to follow a specific structure including Summary, Analysis, Actions, Results, and a narrative Story Explanation. This ensures consistency and makes the output compatible with voice systems.

Can CORE handle parallel tasks?

Yes, CORE includes a delegation framework that can launch multiple 'intern' agents in parallel. It intelligently selects models like Opus, Sonnet, or Haiku based on the task's complexity and speed requirements.

PAI Core Infrastructure

Name: PAI Core Infrastructure
Author: jeffh

byjeffh

•

生産性とワークフロー

Manages the fundamental identity, response formatting, and operational workflows of your Personal AI Infrastructure.

CORE serves as the central nervous system for the PAI (Personal AI Infrastructure), designed to auto-load at the start of every session to establish a consistent personality and operational framework. It enforces a mandatory, structured response format optimized for both readability and voice output, while managing complex workflow routing for Git, CLI, and task delegation. By centralizing stack preferences, security protocols, and historical context retrieval, CORE ensures that Claude Code operates as a highly specialized, secure, and context-aware extension of the user's development environment, capable of parallelizing tasks across multiple model tiers for maximum efficiency.

主な機能

01Intelligent Workflow Routing (Auto-directs tasks for Git, CLI, and merge operations)

028 GitHub stars

03Integrated History System (Fast retrieval of past session learnings and research data)

04Parallel Agent Delegation (Orchestrates sub-agents using Opus, Sonnet, or Haiku models)

05Standardized Response Formatting (MANDATORY format for readability and voice integration)

06Security & Protocol Enforcement (Strict separation of private/public data and prompt injection defense)

ユースケース

01Parallelizing high-volume research or coding tasks by deploying multiple sub-agents simultaneously.

02Automating secure Git workflows and CLI tool creation using predefined system principles.

03Establishing a consistent AI persona and structured communication pattern for all development tasks.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jeffh/claude-plugins CORE

For use in Claude.ai and ChatGPT

Download Skill