@idea - Requirements Gathering

Deep interviewing to capture comprehensive feature requirements. Creates Beads task with PRODUCT_VISION alignment.

When to Use

•Starting new feature
•Unclear requirements
•Need comprehensive spec with tradeoffs explored

Invocation

bash

@idea "feature description"
@idea "feature description" --spec path/to/SPEC.md  # with existing spec

Workflow

Step 1: Read Context

bash

Read(PRODUCT_VISION.md)  # Align with project goals
Glob("docs/specs/**/*")   # Similar features

Step 2: Vision Interview (AskUserQuestion)

Ask product-level questions first:

markdown

AskUserQuestion({
  "questions": [
    {"question": "What is the core mission of this feature?", "header": "Mission",
     "options": [
       {"label": "Solve pain point", "description": "Addresses user frustration"},
       {"label": "Enable capability", "description": "Unlocks new possibility"},
       {"label": "Improve efficiency", "description": "Faster/cheaper process"}
     ]},
    {"question": "How does this align with PRODUCT_VISION.md?", "header": "Alignment",
     "options": [
       {"label": "Core to mission", "description": "Directly supports vision"},
       {"label": "Enables mission", "description": "Supporting capability"},
       {"label": "New direction", "description": "May need vision update"}
     ]}
  ]
})

Step 3: Deep Dive Interview

Continue with technical questions until complete:

Topics to cover:

•Problem & users
•Technical approach (storage, failure modes)
•UI/UX specifics
•Security & performance
•Testing strategy
•Edge cases & tradeoffs

DON'T ask obvious questions. Instead ask about:

•Ambiguities and hidden assumptions
•Tradeoffs between approaches
•Failure modes and edge cases

Step 4: Create Beads Task

python

from sdp.beads import create_beads_client, BeadsTaskCreate, BeadsPriority

client = create_beads_client()

task = client.create_task(BeadsTaskCreate(
    title=feature_title,
    description=f"""## Context & Problem
{problem_answer}

## Goals & Non-Goals
{goals_answer}

## Technical Approach
{technical_answer}

## Concerns & Tradeoffs
{concerns_answer}
""",
    priority=BeadsPriority.HIGH,
    sdp_metadata={
        "feature_type": "idea",
        "mission": mission_answer,
        "product_vision_alignment": alignment_answer,
    }
))

print(f"✅ Created: {task.id}")

Step 5: Create Intent File

python

import json

with open(f"docs/intent/{task.id}.json", "w") as f:
    json.dump({
        "task_id": task.id,
        "title": task.title,
        "mission": mission_answer,
        "alignment": alignment_answer,
        "interview_answers": all_answers,
    }, f, indent=2)

Output

Primary: Beads task ID (e.g., bd-0001)

Secondary:

•docs/intent/{task_id}.json — Machine-readable intent
•Optional: docs/drafts/beads-{task_id}.md — Human-readable export

Next Steps

bash

@design bd-0001      # Decompose into workstreams
bd show bd-0001      # View task details
bd ready             # Check ready tasks

Key Principles

•Start broad, go deep — foundational questions first
•Product vision first — align before technical details
•No obvious questions — explore tradeoffs, not yes/no
•Continue until complete — no ambiguities remain

Example Session

bash

@idea "Add user authentication"

# Vision interview...
# Technical interview...

✅ Created Beads task: bd-0001
   Title: Add user authentication
   Priority: HIGH
   Intent: docs/intent/bd-0001.json

# Next:
@design bd-0001

Quick Reference

Command	Purpose
`@idea "feature"`	Create task with requirements
`@think "complex"`	Deep analysis before @idea
`bd show {id}`	View task details
`@design {id}`	Decompose into workstreams

Version: 3.0.0
See Also: @design, @build, @think