next

Suggest Next Feature

Analyze all features and recommend which one to work on next based on priority, dependencies, and completion status.

Your Task

Evaluate all features in the features/ directory and provide a prioritized recommendation for which feature to implement next.

Process

Step 1: Discover All Features

Use Glob to find all feature directories:

Glob pattern: features/*/context.md

Extract feature names from the paths.

Step 2: Analyze Each Feature

For each discovered feature, gather:

2.1 Completion Status

Read features/[feature-name]/tasks.md and count:

• Total tasks: Match pattern ## Task \d+: or ### Task \d+:
• Completed tasks: Match pattern \[IMPLEMENTED\] or \[COMPLETE\]
• Calculate percentage: completed / total * 100

Categorize status:

Percentage	Status
0%	Not Started
1-99%	In Progress
100%	Complete

2.2 Priority Indicators

From naming convention:

• Extract numeric prefix (e.g., 01-feature-name → priority 1)
• Features without prefix default to priority 99

From context.md:

• Look for "Priority:" or "priority:" mentions
• Look for keywords: "critical", "urgent", "blocker", "MVP", "required"

2.3 Dependencies

Read features/[feature-name]/context.md Dependencies section.

Look for:

• Explicit mentions of other features
• Phrases like "depends on", "requires", "after", "needs"
• References to other feature names

Build a dependency map:

feature-a:
  depends_on: []
  blocks: [feature-b, feature-c]
feature-b:
  depends_on: [feature-a]
  blocks: [feature-d]

Step 3: Calculate Recommendation Scores

For each incomplete feature (status != Complete), calculate a score:

score = (priority_score * 0.40) + (dependency_score * 0.35) + (progress_score * 0.25)

Priority Score (40% weight)

priority_score = 100 - (priority_number * 5)

• Priority 1 → 95 points
• Priority 2 → 90 points
• Priority 10 → 50 points
• Priority 99 (no prefix) → 5 points