Nixtla PRD to Code

Transform Product Requirements Documents into comprehensive implementation task lists with automatic TodoWrite integration for seamless development planning.

Overview

This skill bridges the gap between requirements and code:

• Parse PRDs: Extract functional requirements, non-functional requirements, and technical specifications
• Identify dependencies: Detect task dependencies and ordering constraints
• Generate task lists: Create detailed, prioritized implementation tasks
• TodoWrite integration: Automatically populate Claude's todo list
• Track progress: Maintain clear roadmap from idea to working code

Prerequisites

Required:

• Python 3.8+
• PRD documents in standardized format (Overview, Functional Requirements, Technical Spec)
• Access to TodoWrite tool in conversation context

Optional:

• pyyaml: For YAML task list export (install via pip install pyyaml)

Instructions

Step 1: Identify PRD Document

Locate the PRD file to transform:

ls 000-docs/000a-planned-plugins/*/02-PRD.md

Step 2: Parse PRD

Execute the PRD parser to extract requirements:

python {baseDir}/scripts/parse_prd.py \
    --prd 000-docs/000a-planned-plugins/implemented/nixtla-roi-calculator/02-PRD.md \
    --output tasks.json

Step 3: Review Generated Tasks

The script generates a structured task list with:

• Task titles and descriptions
• Priority levels (P0, P1, P2)
• Dependencies between tasks
• Estimated complexity
• Implementation notes

Step 4: Populate TodoWrite

Automatically populate Claude's todo list:

python {baseDir}/scripts/parse_prd.py \
    --prd 000-docs/000a-planned-plugins/implemented/nixtla-roi-calculator/02-PRD.md \
    --populate-todo

Step 5: Begin Implementation

Follow the generated task list, marking items complete as work progresses.

Output

• tasks.json: Structured task list in JSON format
• tasks.yaml: Human-readable task list (if pyyaml installed)
• implementation_plan.md: Markdown checklist for manual tracking
• TodoWrite integration: Automatic task population in conversation

Error Handling

1. Error: PRD file not found

Solution: Verify PRD path, check 000-docs/000a-planned-plugins/ directory

1. Error: Missing Functional Requirements section

Solution: Ensure PRD has ## Functional Requirements heading with FR-X items

1. Error: TodoWrite tool not available

Solution: Skill can only populate todos in conversation context, not standalone script execution

1. Error: Invalid PRD format

Solution: PRD must have standard sections: Overview, Goals, Functional Requirements, Technical Spec

1. Error: Circular dependency detected

Solution: Review task dependencies, ensure no circular references (Task A → Task B → Task A)

Examples

Example 1: Parse ROI Calculator PRD

python {baseDir}/scripts/parse_prd.py \
    --prd 000-docs/000a-planned-plugins/implemented/nixtla-roi-calculator/02-PRD.md \
    --output roi_tasks.json \
    --verbose

Generated tasks.json:

{
  "tasks": [
    {
      "id": "roi-001",
      "title": "Implement cost input collection",
      "description": "Build input form for infrastructure costs, forecasting volume, team composition",
      "priority": "P0",
      "dependencies": [],
      "complexity": "medium",
      "functional_requirement": "FR-1"
    },
    {
      "id": "roi-002",
      "title": "Build ROI calculation engine",
      "description": "5-year TCO calculation for build vs. buy scenarios",
      "priority": "P0",
      "dependencies": ["roi-001"],
      "complexity": "high",
      "functional_requirement": "FR-2"
    }
  ]
}

Example 2: Auto-Populate TodoWrite

When used in conversation context:

# In Claude Code conversation
from parse_prd import PRDParser

parser = PRDParser('000-docs/000a-planned-plugins/implemented/nixtla-roi-calculator/02-PRD.md')
tasks = parser.extract_tasks()

# Automatically populates TodoWrite
for task in tasks:
    TodoWrite(content=task['title'], activeForm=f"Working on {task['title']}", status="pending")

Example 3: Generate Markdown Checklist

python {baseDir}/scripts/parse_prd.py \
    --prd 000-docs/000a-planned-plugins/implemented/nixtla-forecast-explainer/02-PRD.md \
    --output-format markdown \
    --output implementation_plan.md

Generated implementation_plan.md:

# Nixtla Forecast Explainer - Implementation Plan

## Phase 1: Core Infrastructure (P0)
- [ ] Set up project structure and dependencies
- [ ] Create MCP server scaffold
- [ ] Implement SHAP explainability integration

## Phase 2: Feature Development (P0)
- [ ] Build feature importance calculation
- [ ] Implement counterfactual analysis
- [ ] Add time-based contribution decomposition

## Phase 3: Visualization (P1)
- [ ] Generate waterfall charts
- [ ] Create interactive dashboards
- [ ] Export to PDF reports

Example 4: Batch Process Multiple PRDs

for prd in 000-docs/000a-planned-plugins/*/02-PRD.md; do
    plugin_name=$(basename $(dirname "$prd"))
    python {baseDir}/scripts/parse_prd.py \
        --prd "$prd" \
        --output "009-temp-data/task-plans/${plugin_name}_tasks.json"
done

Resources

• PRD Standard: 000-docs/000a-planned-plugins/README.md (PRD structure specification)
• TodoWrite Documentation: Use AskUserQuestion to learn about TodoWrite tool capabilities
• Task Management: Beads (bd CLI) for advanced task tracking

Related Skills:

• nixtla-plugin-scaffolder: Generate plugin structure from PRD
• nixtla-demo-generator: Create Jupyter demos for implementation
• nixtla-test-generator: Build test suites from PRD requirements

Scripts:

• {baseDir}/scripts/parse_prd.py: Main PRD parsing and task generation script
• {baseDir}/assets/templates/task_template.json: Task structure template