Documentation Index
Fetch the complete documentation index at: https://mintlify.com/swt-labs/vibe-better-with-claude-code-vbw/llms.txt
Use this file to discover all available pages before exploring further.
Advanced commands give you direct control over codebase analysis, standalone research, version management, and work profiles.
/vbw:map
Analyze your codebase with parallel Scout teammates and produce structured mapping documents.
/vbw:map [--incremental] [--package=name] [--tier=solo|duo|quad]
How it works
/vbw:map sizes your codebase and selects a Scout tier automatically:
| Tier | Files | Strategy |
|---|
| solo | <200 | Orchestrator maps inline |
| duo | 200–1,000 | 2 Scouts covering combined domains |
| quad | 1,000+ | 4 Scouts, one per domain |
- Tech Stack — languages, frameworks, dependencies
- Architecture — structure, patterns, component boundaries
- Quality — conventions, testing approach, coverage
- Concerns — debt, risks, security surface
Each Scout writes its domain files directly, then the orchestrator synthesizes them into two cross-cutting documents:
INDEX.md — cross-referenced index with validation notes for contradictionsPATTERNS.md — recurring architectural, naming, quality, and dependency patterns
Incremental refresh: If a mapping already exists and fewer than 20% of files have changed, /vbw:map runs in incremental mode — updating only the affected documents.
Monorepo support: Use --package=name to scope the map to a single package within a monorepo.
Security: A PreToolUse hook enforces that Scouts never read .env files or credential files.
Flags
| Flag | Description |
|---|
--incremental | Force incremental refresh |
--package=name | Scope to a single monorepo package |
--tier=solo|duo|quad | Override auto-detected Scout tier |
Example
/vbw:map --package=api --tier=quad
/vbw:init runs /vbw:map automatically for existing codebases. Run /vbw:map directly when you want to refresh the mapping after significant refactoring.
/vbw:research
Run a standalone research task, decoupled from planning.
Use this when you need answers — architecture options, library comparisons, API behavior — before you’re ready to commit to a plan. Research output can inform a later /vbw:vibe --discuss or /vbw:vibe --plan session.
Plugin maintenance
/vbw:whats-new
View changelog entries since your installed version.
/vbw:update
Update VBW to the latest version.
Fetches the latest version and refreshes the plugin cache automatically.
/vbw:uninstall
Remove VBW cleanly.
Removes the statusline configuration, VBW settings from settings.json, and project data in .vbw-planning/. Your source code is untouched.
Work profiles
Profiles combine effort, autonomy, and verification settings into a single switch. Apply a profile with /vbw:profile [name].
| Profile | Effort | Autonomy | Verification |
|---|
default | balanced | standard | standard |
prototype | fast | confident | quick |
production | thorough | cautious | deep |
yolo | turbo | pure-vibe | skip |
- default — balanced throughput with standard confirmation gates. Good for most projects.
- prototype — fast iteration with fewer prompts. Good for exploratory work or demos.
- production — thorough planning, cautious execution, deep QA. Good for high-stakes deployments.
- yolo — maximum speed, no confirmations, QA skipped. For when you know what you’re doing and want zero ceremony.
Switch profiles:
Create a custom profile: Run /vbw:profile without arguments or use /vbw:config to adjust individual settings — VBW detects the drift from your current profile and offers to save it as a new named profile.
Profiles affect all subsequent /vbw:vibe runs until you switch again. Override per-run with --effort=<level> on any lifecycle command.