Claude Code

LNAI exports unified configuration to Claude Code’s native .claude/ format.

Output Structure

.claude/
├── CLAUDE.md          # Symlink → ../.ai/AGENTS.md
├── rules/             # Symlink → ../.ai/rules/
├── skills/<name>/     # Symlinks → ../../.ai/skills/<name>/
├── settings.json      # Generated
└── <overrides>        # Symlinks from .ai/.claude/

File Mapping

Source	Output	Type
`.ai/AGENTS.md`	`.claude/CLAUDE.md`	Symlink
`.ai/rules/`	`.claude/rules/`	Symlink
`.ai/skills/<name>/`	`.claude/skills/<name>/`	Symlink
`.ai/settings.json`	`.claude/settings.json`	Generated
`.ai/.claude/<path>`	`.claude/<path>`	Symlink

Generated settings.json

Settings are passed through directly in Claude format:

{
  "permissions": {
    "allow": ["Bash(git:*)"],
    "deny": ["Read(.env)"]
  },
  "mcpServers": {
    "memory": {
      "command": "npx",
      "args": ["-y", "@anthropic/mcp-server-memory"]
    }
  }
}

Any settings in overrides.claudeCode are deep-merged into the output.