Find relevant files fast with parallel search agents

Quick Start

/scout find campaign-related components

What happens (under 30 seconds):

1. Spawns 3 parallel scout agents
2. Divides codebase intelligently
3. Each agent searches assigned directories
4. Reports findings to main agent
5. Generates consolidated report

Output: plans/reports/scout-251229-{slug}.md

Syntax

/scout [user-prompt] [scale]

Arguments

Examples

Example 1: Find Campaign Files

Input:

/scout find campaign-related files

Process:

Spawning 3 scout agents in parallel...

Agent 1: Searching app/, components/
Agent 2: Searching lib/, utils/
Agent 3: Searching types/, api/

[30 seconds later]

✓ Scout report generated

Found 12 campaign-related files:
- lib/campaign/manager.ts
- lib/campaign/scheduler.ts
- components/CampaignCard.tsx
- components/CampaignList.tsx
- app/api/campaigns/route.ts
- types/campaign.ts
[... 6 more files]

Saved: plans/reports/scout-251229-2145-campaign-files.md

Example 2: Scale Up Search

Input:

/scout authentication implementation 5

Spawns 5 agents for faster, more thorough search in large codebases.

Workflow Integration

# Before planning
/scout find email-related code
/plan implement email automation

# Before implementation
/cook add payment feature
# (internally runs /scout automatically)

# Manual search
/scout find all database models

Performance

• Speed: 30 seconds avg (vs 2-3 min manual)
• Parallelization: 3-10 agents
• Timeout: 3 min per agent
• Token efficiency: Minimal context usage

Related Commands

• /plan - Uses /scout internally
• /cook - Uses /scout internally and implements after scouting

Find files fast. Parallel search agents map your codebase.