/ideate

Generate improvement ideas by deploying multiple domain experts in parallel to analyze your codebase from different perspectives (security, performance, code quality, UX, testing, API/Architecture).

Quick Start

Copy/agileflow:ideate SCOPE=all DEPTH=quick OUTPUT=report

Parameters

Examples

Analyze All Domains - Quick Scope

Copy/agileflow:ideate SCOPE=all DEPTH=quick OUTPUT=report

Deploys 6 experts (security, performance, refactor, ui, testing, api) to generate 3 improvement ideas each. Results in ideation report with high/medium confidence ideas.

Deep Security Analysis

Copy/agileflow:ideate SCOPE=security DEPTH=deep OUTPUT=stories

Focuses on security with deep analysis (5 ideas per expert). Creates stories for high-confidence items immediately.

Performance Analysis with Both Report and Stories

Copy/agileflow:ideate SCOPE=perf DEPTH=deep OUTPUT=both

Analyzes performance with experts (performance, database, api), generates report and creates implementation stories.

Scope Mapping

Output Structure

Ideation Report (output.md)

The command generates docs/08-project/ideation-<YYYYMMDD>.md with:

1. Summary: Total ideas, confidence breakdown
2. High-Confidence Improvements: Ideas agreed by 2+ experts (prioritize these)
3. Medium-Confidence Opportunities: Single expert with evidence (worth exploring)
4. Summary by Category: Table of idea counts per domain
5. Recommended Next Steps: Implementation guidance

Confidence Scoring

Sample Report Section

Copy# Ideation Report
 
**Generated**: 2025-12-30
**Scope**: all
**Depth**: quick
**Total Ideas**: 15 (High: 4, Medium: 8, Low: 3)
 
## 🎯 High-Confidence Improvements
*Agreed by multiple experts - prioritize these*
 
### 1. Add request rate limiting to API
**Category**: Security | **Impact**: High | **Effort**: 1-2 days
**Experts**: Security, API
**Why**: Prevents DDoS attacks and brute force attempts
**Files**: `src/api/middleware/auth.ts`, `src/api/routes/`
 
## 💡 Medium-Confidence Opportunities
*Single expert with evidence - worth exploring*
 
### 5. Implement caching layer for database queries
**Category**: Performance | **Impact**: High | **Effort**: 2-3 days
**Expert**: Performance
**Why**: Your homepage renders 200+ components, this would fix 2-second load time
**Files**: `src/db/queries.ts`, `src/cache/`

Generated Stories (Optional)

When OUTPUT=stories or OUTPUT=both, creates implementation stories in docs/06-stories/ for each high-confidence idea:

Copy---
story_id: US-XXXX
type: improvement
ideation_source: ideation-20251230.md
estimate: 1d
---
 
# US-XXXX: Idea from Ideation
 
## Background
Identified in ideation report. Agreed by: Security, API experts.
 
## Acceptance Criteria
- [ ] Specific criterion 1
- [ ] Specific criterion 2
- [ ] Tests pass
 
## Technical Notes
Implementation approach from ideation analysis.

How It Works

1. Parse Arguments: Determines which experts to deploy and analysis depth
2. Deploy Experts in Parallel: Launches all experts simultaneously
3. Each Expert Generates Ideas: 3-5 specific, actionable improvement ideas
4. Collect Results: Waits for all experts to complete
5. Synthesize: Groups overlapping ideas, scores confidence
6. Generate Output: Creates report (and optional stories)

Best Practices

Get the Most from /ideate

Do This:

• Use SCOPE=all quarterly for comprehensive analysis
• Use specific SCOPE for focused improvement areas
• Use DEPTH=deep when investigating specific domain
• Review HIGH-confidence ideas first
• Create stories for HIGH-confidence improvements

Don't Do This:

• Ignore MEDIUM-confidence ideas with specific evidence
• Try to implement everything at once
• Use without running git to get latest code context
• Skip the "why" section when deciding to implement

Interpreting Results

High-Confidence Ideas (2+ experts):

• Strong signal of real improvement opportunity
• Recommended for immediate action
• Usually require 1-3 days of effort

Medium-Confidence Ideas (single expert):

• Valuable but less certain
• Worth discussing with team
• Often good candidates for next sprint

Vague Ideas:

• Automatically excluded from report
• Too generic without file paths or metrics
• If interested, do focused research first

Related Commands

• /epic - Create epic for ideation improvements
• /story - Create story for specific improvement
• /research:analyze - Deep dive on specific improvement area
• /babysit - Get help implementing an improvement

Integration

With Sprint Planning

After ideation, use high-confidence items to populate sprint:

Copy/agileflow:ideate SCOPE=all DEPTH=quick OUTPUT=stories
# Stories created → add to sprint planning with /agileflow:sprint

With Retrospectives

Run ideation after retrospectives to find systemic improvements:

Copy/agileflow:ideate SCOPE=code DEPTH=deep
# Analyze code quality problems identified in retro

With ADRs

Convert architectural ideas into ADRs:

Copy/agileflow:ideate SCOPE=architecture OUTPUT=report
# Review HIGH-confidence architectural improvements
/agileflow:adr NUMBER=0045 TITLE="..." CONTEXT="From ideation"