knowledge-reorg

To reorganize AGENTS.md, this Skill analyzes repository knowledge and applies the AI Agent Knowledge Architecture Policy to separate entry points, documentation, and executable workflows. It centralizes agent references into a single AGENTS.md hub while extracting documents into dedicated skills.

The three-layer structure for multi-agent knowledge architecture separates entry points, documentation, and executable workflows. This Skill enforces that structure by compressing entrypoints, extracting documents, and checking link integrity across Claude Code, Codex CLI, Kiro CLI, and Gemini CLI setups.

Migrating agent knowledge architecture across Claude Code, Codex CLI, Kiro CLI, and Gemini CLI requires centralizing entrypoints into a single AGENTS.md hub and separating docs from skills. This Skill provides guided migration with cross-agent imports and consistency checks for multi-agent integration.

Yes, this knowledge architecture reorganization supports optional Serena memory migration. Alongside entrypoint compression and document extraction, it includes Serena memory migration to maintain knowledge continuity when reorganizing AGENTS.md and Skills per policy.

You should refactor your repository's AGENTS.md and skills for multi-agent setups when knowledge discoverability and maintainability decline across agents. This Skill applies the AI Agent Knowledge Architecture Policy to separate concerns and improve cross-agent consistency.

The best way to compress AGENTS.md entrypoints without breaking links is to run a guided reorganization workflow that performs entrypoint compression, document extraction, and link integrity checks. This ensures the three-layer structure is satisfied while maintaining cross-agent references.