> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/derHaken/SuperAntigravity/llms.txt
> Use this file to discover all available pages before exploring further.

# Commands overview

> 16 slash commands for explicit workflow control in Antigravity

SuperAntigravity ships 16 slash commands that give you explicit control over your development workflow. While [skills](/skills/overview) fire automatically when context matches, commands are invoked deliberately — you decide when to run them.

## How to use commands

Type `/` followed by the command name in any Antigravity conversation:

```bash theme={null}
/brainstorm
/plan
/implement
```

Each command is backed by a workflow file in `~/.gemini/antigravity/global_workflows/`. You can inspect and customize any workflow file directly.

## All commands

| Command                                   | What it does                                                 |
| ----------------------------------------- | ------------------------------------------------------------ |
| [`/brainstorm`](/commands/brainstorm)     | \[Step 1] Requirements discovery before writing code         |
| [`/plan`](/commands/plan)                 | \[Step 2] Write a detailed implementation plan               |
| [`/implement`](/commands/implement)       | \[Step 3] Feature implementation with TDD                    |
| [`/review`](/commands/review)             | \[Step 4] Code review against plan and quality standards     |
| [`/git`](/commands/git)                   | \[Step 5] Git operations with smart commits                  |
| [`/research`](/commands/research)         | Deep multi-source research with confidence levels            |
| [`/build`](/commands/build)               | Build and package the project                                |
| [`/test`](/commands/test)                 | Run tests with coverage analysis                             |
| [`/analyze`](/commands/analyze)           | Code quality, security, performance, and architecture review |
| [`/design`](/commands/design)             | System architecture, API, and component design               |
| [`/troubleshoot`](/commands/troubleshoot) | Systematic debugging with phase-by-phase isolation           |
| [`/explain`](/commands/explain)           | Explain code, concepts, or system behavior                   |
| [`/improve`](/commands/improve)           | Apply systematic code improvements                           |
| [`/cleanup`](/commands/cleanup)           | Remove dead code and unused imports                          |
| [`/document`](/commands/document)         | Generate documentation for components and APIs               |
| [`/estimate`](/commands/estimate)         | Development estimates with confidence levels                 |

### Bonus commands

These commands are available but not part of the core workflow:

| Command           | What it does                                                  |
| ----------------- | ------------------------------------------------------------- |
| `/business-panel` | AI-facilitated panel discussion with business thought leaders |
| `/spec-panel`     | Multi-expert review of technical specification quality        |
| `/recommend`      | Intelligent recommender for SuperAntigravity commands         |

## Conversation mode guidance

Antigravity's conversation modes affect how commands perform. Match the mode to the command for best results.

<Warning>
  **Planning Mode** is mandatory for `/brainstorm`, `/plan`, `/research`, and `/troubleshoot`. These commands require deep thinking, design reasoning, and complex problem isolation — they will not perform correctly in Fast Mode.
</Warning>

<Tip>
  **Fast Mode** is recommended for `/implement`, `/test`, and batch execution. These commands are optimized for direct code generation and mechanical tasks.
</Tip>

| Mode                     | Commands                                                                                               |
| ------------------------ | ------------------------------------------------------------------------------------------------------ |
| Planning Mode (required) | `/brainstorm`, `/plan`, `/research`, `/troubleshoot`                                                   |
| Fast Mode (recommended)  | `/implement`, `/test`, `/build`                                                                        |
| Either mode              | `/git`, `/analyze`, `/design`, `/explain`, `/improve`, `/review`, `/cleanup`, `/document`, `/estimate` |

## The core workflow

The five numbered commands form a complete development pipeline:

```
/brainstorm  →  discuss + design (no code yet)
     ↓
/plan        →  bite-sized tasks with TDD steps
     ↓
/implement   →  subagent per task + code review between
     ↓
/review      →  validate against plan
     ↓
/git         →  smart commit and push
```

<CardGroup cols={2}>
  <Card title="/brainstorm" href="/commands/brainstorm">
    Start here. Turn vague ideas into approved designs before writing a line of code.
  </Card>

  <Card title="/plan" href="/commands/plan">
    Convert an approved design into a bite-sized, TDD-ready implementation plan.
  </Card>

  <Card title="/implement" href="/commands/implement">
    Execute the plan task by task with confidence checks and review gates.
  </Card>

  <Card title="/troubleshoot" href="/commands/troubleshoot">
    Systematic four-phase debugging when something breaks.
  </Card>
</CardGroup>