claude-rig-awarenesslisted

# claude-rig Awareness claude-rig lets users run multiple independent Claude Code configurations side by side. Each configuration is called a **rig**. When a rig is active, Claude Code doesn't use `~/.claude/` directly — it uses the rig's directory instead. ## How to detect the active rig Check the `CLAUDE_CONFIG_DIR` environment variable. If set, you're running inside a rig. ``` echo $CLAUDE_CONFIG_DIR # Example: /home/user/.claude-rig/rigs/go ``` **If `CLAUDE_CONFIG_DIR` is set, that directory replaces `~/.claude/` for everything.** Settings, plugins, skills, hooks, agents, commands, CLAUDE.md — all of it lives there, not in `~/.claude/`. If the variable is unset, there's no active rig and `~/.claude/` is the config dir as usual. ## The 3-layer config model Configuration resolves in three layers, most specific wins: | Layer | Location | Scope | |---------|----------------------------------------------------|------------------------| | Global | `~/.claude/` | Shared across all rigs | | Rig | `~/.claude-rig/rigs/<name>/` (`CLAUDE_CONFIG_DIR`) | Per-rig override | | Project | `.claude/` in the project working dir | Per-project override | Rig-specific items (settings, plugins, skills, hooks, agents) are real files that override the global ones. New rigs also isolate per-rig data by default (conversations, history, sessions, etc.). Remaining shar