---
description: Get or set AgentVibes verbosity level (low/medium/high/caveman)
tags: [user]
---

Get or set the AgentVibes verbosity level to control how much Claude speaks while working.

## Usage

- `/agent-vibes:verbosity` - Show current verbosity level
- `/agent-vibes:verbosity low` - Set to LOW (minimal)
- `/agent-vibes:verbosity medium` - Set to MEDIUM (balanced)
- `/agent-vibes:verbosity high` - Set to HIGH (maximum transparency)
- `/agent-vibes:verbosity caveman` - Set to CAVEMAN (ultra-terse, max token savings)

## Verbosity Levels

### LOW (Minimal)
- ✅ Acknowledgments only (start of task)
- ✅ Completions only (end of task)
- 🔇 No reasoning spoken during work
- Perfect for: Quiet work sessions, minimal distraction

### MEDIUM (Balanced)
- ✅ Acknowledgments
- 🤔 Major decisions ("I'll use grep to search")
- ✓ Key findings ("Found 12 instances")
- ✅ Completions
- Perfect for: Understanding major decisions without full narration

### HIGH (Maximum Transparency)
- ✅ Acknowledgments
- 💭 All reasoning ("Let me search for all instances")
- 🤔 All decisions ("I'll use grep for this")
- ✓ All findings ("Found it at line 1323")
- ✅ Completions
- Perfect for: Full transparency, learning mode, debugging complex tasks

### CAVEMAN (Ultra-Terse)
- ⚡ Fragments only — no articles, filler, or hedging
- 🔤 Abbreviations (DB/auth/config/fn/impl)
- ➡️ Arrows instead of prose (X -> Y)
- 📉 65-75% fewer output tokens
- Perfect for: Maximum token savings, cost-conscious sessions
- Inspired by [caveman](https://github.com/JuliusBrussee/caveman) (MIT)

## How It Works

Claude uses **emoji markers** in its reasoning text:
- 💭 = Reasoning/thinking
- 🤔 = Decisions
- ✓ = Findings/results

Based on your verbosity level, AgentVibes automatically detects these markers and speaks them aloud.

## Examples

### LOW Verbosity
```
User: "Find all TODO comments"
[TTS: "I'll search for those"]
[Work happens silently...]
[TTS: "Found 12 TODO comments"]
```

### MEDIUM Verbosity
```
User: "Find all TODO comments"
[TTS: "I'll search for those"]
🤔 I'll use grep to search all files
[Work happens...]
✓ Found 12 TODO comments across 5 files
[TTS: "Found 12 TODO comments"]
```

### HIGH Verbosity
```
User: "Find all TODO comments"
[TTS: "I'll search for those"]
💭 Let me search through the codebase for TODO comments
🤔 I'll use the Grep tool with pattern "TODO"
[Grep runs...]
✓ Found 12 TODO comments across 5 files
💭 Let me organize these by file
[TTS: "Found 12 TODO comments in 5 files"]
```

## Notes

- Changes take effect on next Claude Code session restart
- Verbosity is saved per-project (if `.claude/` exists) or globally
- You can also control verbosity via MCP: `mcp__agentvibes__set_verbosity(level="high")`

## Related Commands

- `/agent-vibes:personality` - Set voice personality style
- `/agent-vibes:switch` - Change voice
- `/agent-vibes:provider` - Switch TTS provider