Project Estimation Guide

Overview

This skill helps create accurate software development estimates by systematically breaking down projects into estimable tasks and assigning effort by role. Output can be delivered as Excel spreadsheets using the xlsx skill, and tasks can be pushed directly to Jira to create a project backlog.

Estimation Workflow

Step 1: Clarify Project Scope

ALWAYS ask the user about project maturity level before estimating:

Level	Description	Multiplier
PoC	Proof of Concept - minimal viable functionality, no production concerns	0.3-0.5x
MVP	Minimum Viable Product - core features, basic quality, limited scale	0.6-0.8x
Production	Full production - complete features, security, scalability, monitoring	1.0x

Key questions to ask:

• Is this PoC, MVP, or production-ready system?
• What are the non-functional requirements (users, load, compliance)?
• Are there existing systems to integrate with?
• What is the team's familiarity with the tech stack?
• Are there hard deadlines or constraints?

Step 2: Hierarchical Task Breakdown

Break down from high-level to lowest estimable units (1-3 days max per task).

Level 0: Project
└── Level 1: Epic/Module
    └── Level 2: Feature
        └── Level 3: Task
            └── Level 4: Subtask (if needed)

Example breakdown:

E-commerce Platform
├── User Management (Epic)
│   ├── Registration (Feature)
│   │   ├── Registration form UI (Task) - FE: 1.5d
│   │   ├── Email verification API (Task) - BE: 1d
│   │   ├── Email templates (Task) - BE: 0.5d
│   │   └── Registration tests (Task) - QA: 1d
│   ├── Authentication (Feature)
│   │   ├── Login/logout UI (Task) - FE: 1d
│   │   ├── JWT auth endpoints (Task) - BE: 1.5d
│   │   ├── Password reset flow (Task) - BE: 1.5d, FE: 1d
│   │   └── Auth tests (Task) - QA: 1.5d
│   └── User Profile (Feature)
│       └── ...
├── Product Catalog (Epic)
│   └── ...
└── Checkout (Epic)
    └── ...

Step 3: Assign Effort by Role

Standard roles and responsibilities:

Role	Code	Responsibilities
Backend Engineer	BE	APIs, business logic, database, integrations
Frontend Engineer	FE	UI components, state management, UX implementation
AI/ML Engineer	AI	Model development, training, inference pipelines
QA Engineer	QA	Test planning, automation, manual testing
DevOps Engineer	DevOps	Infrastructure, CI/CD, monitoring, deployment
Project Manager	PM	Coordination, documentation, stakeholder communication
Designer	Design	UI/UX design, prototypes, design system
Tech Lead	TL	Architecture, code review, technical decisions

Step 4: Apply Risk Factors

Risk Factor	Multiplier
New technology for team	1.3-1.5x
Unclear requirements	1.2-1.5x
External API dependencies	1.2-1.4x
Legacy system integration	1.3-1.6x
Tight deadline	1.2-1.4x
Distributed team	1.1-1.3x

Step 5: Add Contingency Buffer

Confidence Level	Buffer
High (clear scope, known tech)	10-15%
Medium (some unknowns)	20-30%
Low (many unknowns)	30-50%

Baseline Estimates by Component

Frontend Components

Component	FE Effort	Notes
Simple page	0.5-1d	Static content
Form (simple)	1-2d	3-5 fields, basic validation
Form (complex)	3-5d	Multi-step, conditional logic
Data table	2-4d	Sort, filter, pagination
Dashboard	3-5d	Multiple widgets
Admin panel	5-10d	CRUD operations, filters

Backend Components

Component	BE Effort	Notes
CRUD endpoint	0.5-1d	Standard operations
Complex endpoint	1-3d	Business logic, validation
Third-party integration	2-5d	Per API
Background job	1-3d	With retry logic
Auth system	3-5d	JWT, sessions, permissions

AI/ML Components

Component	AI Effort	Notes
LLM integration	2-4d	Prompt engineering, API setup
RAG pipeline	5-10d	Embeddings, vector DB, retrieval
Custom model training	10-20d	Data prep, training, evaluation
ML inference API	3-5d	Model serving, optimization

QA Effort Guidelines

Coverage	QA Multiplier
Minimal (smoke tests)	0.1x of dev
Standard (core flows)	0.2-0.3x of dev
Comprehensive	0.4-0.5x of dev
Critical system	0.5-0.7x of dev

PM/Coordination Overhead

Project Size	PM Overhead
Small (<1 month)	10-15%
Medium (1-3 months)	15-20%
Large (3+ months)	20-25%

Output Format

Excel Spreadsheet (Recommended)

Use the xlsx skill to create estimates in Excel format with:

Sheet 1: Summary

Role	Days	Rate	Cost
BE	45	$X	$Y
FE	30	$X	$Y
QA	15	$X	$Y
Total	90		$Z

Sheet 2: Detailed Breakdown

Epic	Feature	Task	BE	FE	AI	QA	DevOps	PM	Total
User Mgmt	Registration	Form UI	-	1.5	-	-	-	-	1.5
User Mgmt	Registration	API	1	-	-	-	-	-	1
User Mgmt	Registration	Tests	-	-	-	1	-	-	1

Sheet 3: Assumptions & Risks

• Document all assumptions made
• List identified risks with mitigation strategies
• Note excluded items

Markdown Table (Quick Estimates)

## Estimate Summary

| Phase | BE | FE | QA | DevOps | PM | Total |
|-------|----|----|----|----|-----|-------|
| Phase 1 | 20d | 15d | 8d | 5d | 5d | 53d |
| Phase 2 | 15d | 10d | 6d | 3d | 4d | 38d |
| Buffer (20%) | | | | | | 18d |
| **Total** | **35d** | **25d** | **14d** | **8d** | **9d** | **109d** |

Common Pitfalls to Avoid

Underestimation causes:

• Forgetting testing, code review, deployment time
• Not accounting for meetings (add 10-20%)
• Missing error handling and edge cases
• Ignoring documentation needs
• Assuming ideal conditions

Items often forgotten:

• Authentication and authorization
• Error handling and logging
• Data migration
• Environment setup
• API documentation
• Security hardening
• Performance optimization
• Monitoring and alerting

Quick Reference

T-Shirt to Days Conversion

Size	PoC	MVP	Production
XS	0.25d	0.5d	1d
S	0.5d	1d	2d
M	1d	2-3d	4-5d
L	2d	4-5d	8-10d
XL	3-5d	8-10d	15-20d

Story Points to Days (per developer)

Points	Days
1	0.25-0.5
2	0.5-1
3	1-2
5	2-3
8	3-5
13	5-8

Jira Integration

Push estimated tasks directly to Jira to create a project backlog.

Prerequisites

Before pushing to Jira, gather from the user:

1. Jira instance URL (e.g., https://company.atlassian.net)
2. Project key (e.g., PROJ)
3. API token - user can generate at https://id.atlassian.com/manage-profile/security/api-tokens
4. Email associated with the Jira account

Task to Jira Mapping

Estimation Level	Jira Issue Type
Epic	Epic
Feature	Story
Task	Task or Sub-task
Subtask	Sub-task

Field Mapping

Estimate Field	Jira Field
Epic name	Epic Name (custom field)
Feature/Task name	Summary
Description	Description
Total effort (days)	Original Estimate (convert to hours: days × 8)
Role assignments	Labels or custom field
Story points	Story Points (if using points)

Push to Jira Workflow

Step 1: Confirm with user

Before pushing to Jira, confirm:
- Jira URL: [user provides]
- Project key: [user provides]
- Create Epics? (yes/no)
- Estimate format: (days/hours/story points)

Step 2: Create issues via API

Use bash with curl to create issues. First, set up authentication:

# Store credentials (ask user for these values)
JIRA_URL="https://company.atlassian.net"
JIRA_EMAIL="user@company.com"
JIRA_TOKEN="your-api-token"
PROJECT_KEY="PROJ"

Create an Epic:

curl -X POST "$JIRA_URL/rest/api/3/issue" \
  -H "Content-Type: application/json" \
  -u "$JIRA_EMAIL:$JIRA_TOKEN" \
  -d '{
    "fields": {
      "project": {"key": "'"$PROJECT_KEY"'"},
      "summary": "User Management",
      "issuetype": {"name": "Epic"},
      "description": {
        "type": "doc",
        "version": 1,
        "content": [{"type": "paragraph", "content": [{"type": "text", "text": "Epic for user management features"}]}]
      }
    }
  }'

Create a Story linked to Epic:

curl -X POST "$JIRA_URL/rest/api/3/issue" \
  -H "Content-Type: application/json" \
  -u "$JIRA_EMAIL:$JIRA_TOKEN" \
  -d '{
    "fields": {
      "project": {"key": "'"$PROJECT_KEY"'"},
      "summary": "User Registration",
      "issuetype": {"name": "Story"},
      "parent": {"key": "PROJ-1"},
      "description": {
        "type": "doc",
        "version": 1,
        "content": [{"type": "paragraph", "content": [{"type": "text", "text": "Implement user registration flow"}]}]
      },
      "timetracking": {
        "originalEstimate": "24h"
      },
      "labels": ["BE", "FE", "QA"]
    }
  }'

Create a Task/Sub-task:

curl -X POST "$JIRA_URL/rest/api/3/issue" \
  -H "Content-Type: application/json" \
  -u "$JIRA_EMAIL:$JIRA_TOKEN" \
  -d '{
    "fields": {
      "project": {"key": "'"$PROJECT_KEY"'"},
      "summary": "Registration form UI",
      "issuetype": {"name": "Sub-task"},
      "parent": {"key": "PROJ-2"},
      "timetracking": {
        "originalEstimate": "12h"
      },
      "labels": ["FE"]
    }
  }'

Batch Creation Script

For large estimates, generate a bash script that creates all issues:

#!/bin/bash
# Auto-generated Jira import script

JIRA_URL="https://company.atlassian.net"
JIRA_EMAIL="$JIRA_EMAIL"
JIRA_TOKEN="$JIRA_TOKEN"
PROJECT_KEY="PROJ"

create_issue() {
  local json="$1"
  response=$(curl -s -X POST "$JIRA_URL/rest/api/3/issue" \
    -H "Content-Type: application/json" \
    -u "$JIRA_EMAIL:$JIRA_TOKEN" \
    -d "$json")
  echo "$response" | grep -o '"key":"[^"]*"' | cut -d'"' -f4
}

# Create Epics first, capture keys
echo "Creating Epics..."
EPIC_USER_MGMT=$(create_issue '{"fields":{"project":{"key":"'"$PROJECT_KEY"'"},"summary":"User Management","issuetype":{"name":"Epic"}}}')
echo "Created: $EPIC_USER_MGMT"

# Create Stories under Epics
echo "Creating Stories..."
STORY_REGISTRATION=$(create_issue '{"fields":{"project":{"key":"'"$PROJECT_KEY"'"},"summary":"User Registration","issuetype":{"name":"Story"},"parent":{"key":"'"$EPIC_USER_MGMT"'"},"timetracking":{"originalEstimate":"24h"}}}')
echo "Created: $STORY_REGISTRATION"

# Create Sub-tasks under Stories
echo "Creating Tasks..."
create_issue '{"fields":{"project":{"key":"'"$PROJECT_KEY"'"},"summary":"Registration form UI","issuetype":{"name":"Sub-task"},"parent":{"key":"'"$STORY_REGISTRATION"'"},"timetracking":{"originalEstimate":"12h"},"labels":["FE"]}}'

echo "Done!"

CSV Export for Jira Import

Alternatively, create a CSV file for Jira's CSV import feature:

Summary,Issue Type,Parent,Description,Original Estimate,Labels
User Management,Epic,,"Epic for user management features",,
User Registration,Story,User Management,"Implement user registration",24h,"BE,FE,QA"
Registration form UI,Sub-task,User Registration,"Build the registration form",12h,FE
Email verification API,Sub-task,User Registration,"API endpoint for email verification",8h,BE
Registration tests,Sub-task,User Registration,"Write tests for registration",8h,QA

CSV Import Steps:

1. Go to Jira > Project Settings > External System Import
2. Select "CSV"
3. Upload the CSV file
4. Map columns to Jira fields
5. Review and import

Estimate to Hours Conversion

When pushing to Jira, convert days to hours:

Days	Hours (8h/day)
0.25	2h
0.5	4h
1	8h
1.5	12h
2	16h
3	24h
5	40h

Jira Custom Fields

Common custom fields that may need configuration:

Field	API Field Name	Notes
Story Points	customfield_10016	ID varies by instance
Epic Link	customfield_10014	For older Jira versions
Sprint	customfield_10020	ID varies by instance

To find custom field IDs for a project:

curl -s "$JIRA_URL/rest/api/3/field" \
  -u "$JIRA_EMAIL:$JIRA_TOKEN" | jq '.[] | {name, id}'

Error Handling

Common issues when pushing to Jira:

Error	Cause	Solution
401 Unauthorized	Invalid credentials	Verify email and API token
400 Bad Request	Invalid field values	Check issue type names, project key
403 Forbidden	Insufficient permissions	User needs project write access
Field not found	Custom field ID wrong	Look up correct field ID

Jira Cloud vs Server

The API examples above use Jira Cloud (Atlassian Document Format for descriptions). For Jira Server/Data Center:

# Jira Server uses plain text descriptions
curl -X POST "$JIRA_URL/rest/api/2/issue" \
  -H "Content-Type: application/json" \
  -u "$JIRA_EMAIL:$JIRA_TOKEN" \
  -d '{
    "fields": {
      "project": {"key": "'"$PROJECT_KEY"'"},
      "summary": "Task name",
      "issuetype": {"name": "Task"},
      "description": "Plain text description here"
    }
  }'