> ## Documentation Index
> Fetch the complete documentation index at: https://docs.launchpulse.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Agents

> Understand LaunchPulse build modes and choose the right agent workflow for your product stage.

# Agents

LaunchPulse agents are your build system: they transform product intent into scoped plans, implementation steps, and quality loops. Choosing the right mode keeps your team faster and your outputs more predictable.

## How the LaunchPulse build system works

<Steps>
  <Step title="Plan">
    Your prompt is converted into a structured feature plan.
  </Step>

  <Step title="Prioritise">
    You remove, reorder, and clarify scope before execution.
  </Step>

  <Step title="Build">
    Agent workflows execute implementation based on approved priorities.
  </Step>

  <Step title="Validate">
    Testing and review agents help catch gaps before launch.
  </Step>
</Steps>

## Agent comparison

| Option                          | Best for                          | How it works                                         | When to start with it                     |
| ------------------------------- | --------------------------------- | ---------------------------------------------------- | ----------------------------------------- |
| Autonomous AI Software Engineer | First MVP builds                  | One primary agent handles plan-to-build flow         | Day 1 for most founders                   |
| Feature Builder                 | Better scoping before coding      | Generates editable feature plan and priorities       | Before approving a build                  |
| Custom Agents                   | Repeated specialist tasks         | Reusable role-based agents with custom prompts/tools | When workflows repeat                     |
| AI Multi-Agent Mode             | Larger or complex projects        | Specialist agents collaborate via handoffs           | When work needs role separation           |
| Testing Agent                   | Higher release confidence         | Structured test planning and validation loops        | Before major launches                     |
| LaunchPulseClaw                 | Early advanced extension workflow | Hosted OpenClaw-based extension path                 | If you need experimental advanced control |

## Which one should I use?

* **I'm launching first MVP quickly:** start with **Autonomous AI Software Engineer**
* **My prompts are broad and messy:** add **Feature Builder** before build approval
* **My team repeats QA/review prompts:** create **Custom Agents**
* **My product has many modules/owners:** explore **AI Multi-Agent Mode**
* **I'm preparing for launch week:** run **Testing Agent**

## Example scenarios

* Solo founder shipping a B2B SaaS MVP in one focused sprint
* Product team splitting planning, implementation, and QA responsibilities
* Agency reusing custom review/testing agent setups across client projects

## Screenshot placeholders

> \[Screenshot placeholder: Agent selection screen during project setup]

> \[Screenshot placeholder: Multi-agent workflow or handoff view]

## Related pages

<CardGroup cols={3}>
  <Card title="Autonomous AI Software Engineer" href="/autonomous-ai-software-engineer" />

  <Card title="Feature Builder" href="/feature-builder" />

  <Card title="Custom Agents" href="/custom-agents" />

  <Card title="AI Multi-Agent Mode" href="/ai-multi-agent-mode" />

  <Card title="Testing Agent" href="/testing-agent" />

  <Card title="LaunchPulseClaw" href="/launchpulseclaw" />
</CardGroup>