import-host

Detect existing CLAUDE Code and agent files and propose pin updates to USER.md and PROJECT.md.

Updated Apr 24, 2026
One-click install
npx skills add https://github.com/Cain-Ish/claude-code-plugin --skill import-host
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: import-host
Source: https://github.com/Cain-Ish/claude-code-plugin/tree/main/skills/import-host
Command: npx skills add https://github.com/Cain-Ish/claude-code-plugin --skill import-host

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It eliminates the effort of manually copying prior Claude Code and agent-rule knowledge into Second Brain hot-tier files by detecting what already exists on your host.

Core Features & Use Cases

  • Detect host AI-context files and classify their contents as user-level preferences versus project-level facts.
  • Propose safe pins and targeted updates to USER.md and PROJECT.md with explicit confirmations, without deleting or overwriting host sources.
  • Populate hot-tier from common authoring locations like ~/CLAUDE.md, ~/AGENTS.md, and repo-root agent config files to speed up setup.

Quick Start

Run /second-brain:import-host and confirm the proposed USER.md preferences and PROJECT.md updates chunk-by-chunk.

Frequently Asked Questions about import-host

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

FAQPage Schema
How do I import existing Claude Code context files into a local knowledge base?

Importing Claude Code preferences classifies detected context into user-level preferences and project-level facts, proposing safe pins to USER.md and PROJECT.md chunk-by-chunk with explicit confirmation gating to prevent unauthorized writes.

How does importing host AI-context handle duplicate entries during reruns?

Importing host AI-context handles duplicates through idempotent behavior, safely skipping existing entries during bootstrap or reruns to ensure prior user preferences and project facts are not duplicated in the hot-tier memory.

Does importing agent-rule files require write access to my existing repository configs?

Importing agent-rule files does not require write access to existing repository configs; the Skill performs read-only scanning of $HOME and repo roots, only writing idempotent updates to USER.md and PROJECT.md after confirmation.

What is the best way to bootstrap project knowledge from existing agent config files?

The best way to bootstrap project knowledge from agent config files is running an automated discovery scan that detects existing rules, classifies them as user or project facts, and proposes confirmation-gated pins to populate hot-tier files.

Do I need YAML-defined discovery metadata to scan my host AI-context files?

Yes, you need YAML-defined discovery metadata to enable the read-only scanning of $HOME and current repo root, which identifies existing agent-rule files and proposes targeted hot-tier updates for your local wiki.