3 Guides
Claude (Anthropic)
Role-Based Prompting
Agile / Scrum Teams
Free · .docx
What Is Prompt Engineering — and Why Does Your Scrum Team Need It?
Prompt engineering is the discipline of crafting precise, structured instructions for an AI model to produce expert-quality outputs. For Agile teams, it is the difference between getting a generic answer and receiving a ready-to-use professional document.
Role Assignment
Telling the AI "You are the Scrum Master of Team 1" activates domain-specific knowledge. The AI reasons as an expert practitioner — not as a generic assistant — producing context-aware, professional output.
Structured Decomposition
Breaking the request into numbered sub-tasks (1. Team structure, 2. Scrum events, 3. Timeline...) forces the AI to address each requirement systematically. Vague prompts produce vague results.
Context Anchoring
Referencing a specific project ("AI-enabled flight booking system") grounds the AI's output in real-world examples relevant to your team — not generic Scrum theory that teams have already read a hundred times.
Output Specification
Specifying the format ("prepare a Word copy", "include tables", "show a sprint calendar") controls how the AI presents its knowledge. Format matters as much as content for adoption in your team.
The 3 Expert Prompts — Refined & Annotated
These are the exact prompts used to generate the three Agile guides, refined with prompt engineering best practices and annotated so you understand why each element works.
Generates: Team structure · Scrum events · Sprint calendar · Project timeline · Metrics framework
// You also carry the responsibilities of the Project Manager.
// Generate a comprehensive, professional Scrum Master Guide covering:
1. TEAM STRUCTURE — Show a team of exactly 10 members (yourself as SM, a Product Owner, and 8 developers with specialised roles). For each member include: role, name, core responsibility, and AI focus area relevant to flight booking.
2. SCRUM EVENTS — List all Scrum events you will facilitate as SM, including the weekly Backlog Refinement (Grooming) session. For each event provide: frequency, duration, agenda, presenter, attendees, inputs, outputs, and action items to sustain project flow.
3. PROJECT TIMELINE — Show a full delivery plan with: high-level goals for every 3-month quarter, sprint goals for every individual sprint, and key deliverables. Structure the timeline as a PM would — tracking milestones, not just Scrum artefacts.
4. SPRINT CALENDAR — Prepare a complete 2026 sprint calendar where: each sprint starts on a Monday and ends on a Friday, duration is exactly 2 weeks (10 working days), and all 4 quarters are colour-coded. Note any year-end schedule adjustments.
5. SPRINT METRICS — Define a weekly metrics framework covering Delivery, Quality, Flow, Team Health, and AI/Technical categories. For each metric include: definition, how to measure it, target value, frequency, and how to use it in data-driven retrospectives and management reports.
// Format: Prepare a professional Word document with tables, colour-coded sections,
// a header/footer, and section banners. Use Arial font. Write a simple, practical guide.
Why This Prompt Works
The prompt opens with an explicit role assignment (“You are the Scrum Master”) which activates practitioner-level Agile knowledge. Each numbered section specifies exactly what sub-topic to address, eliminating ambiguity. The phrase “You also carry PM responsibilities” is critical — it signals the AI to blend Scrum discipline with project management rigour (timelines, milestones, reporting). The closing format instruction ensures the output is document-ready, not just conversational text.
Key Outputs This Prompt Generates
10-member team roster
6 Scrum event facilitation guides
24-sprint project timeline
2026 sprint calendar
25-metric performance framework
Tips to Personalise This Prompt
Change the project
Adjust team size
Change the year
Add your tools
// Generate a comprehensive, practical Product Owner Guide covering:
1. PRODUCT SCOPE — Write a high-level scope document (maximum 2 pages) covering the complete end-to-end system view. Include: all functional modules, third-party integrations, in-scope vs out-of-scope boundaries, key user personas, and a non-functional requirements summary (performance, security, scalability, availability).
2. REQUIREMENTS HIERARCHY — Create a parent-child hierarchy in this exact order:
– Epics → Capabilities → Features → User Stories
Show a minimum of 8 Epics with child Capabilities and Features.
Map each item to: Functional vs Non-Functional requirement, and User Requirement vs Enabler Requirement.
Also include a Waterfall SRS view (System Requirements Specification) for teams working in hybrid modes.
3. ESTIMATION & PRIORITISATION — Explain the techniques to be used after Epics/Features/Stories are created and refined. Cover: Planning Poker, T-Shirt Sizing, WSJF, MoSCoW, Kano Model, and Value vs Effort Matrix. Show a 6-step Refinement → Estimation → Prioritisation workflow.
4. TEAM BACKLOG ASSIGNMENT — Show how the product backlog is split across 2 Scrum teams. Display team-specific sprint backlogs with: Story ID, description, Feature, Story Points, MoSCoW priority, and Sprint assignment. Include shared backlog governance rules.
5. CAPACITY VS WORKLOAD — Create a capacity tally showing: sprint capacity assumptions, member-level capacity for Team 1 (hours and velocity), and a sprint-by-sprint workload tally for both teams with buffer percentage and traffic-light status (OK / Watch / Tight / Over).
// Format: A clean, professional Word document with colour-coded tables, banners,
// and a cover page. Write a simple, practical guide that a PO can use every sprint.
Why This Prompt Works
Key Outputs This Prompt Generates
2-page product scope document
Epic → Story hierarchy table
5 estimation techniques
Dual-team sprint backlogs
12-sprint capacity tally
Tips to Personalise This Prompt
Specify your Epics
Choose your frameworks
Single team version
Add your velocity
// Generate a comprehensive Scrum Developers Guide for the entire development team.
// The guide must be practical, opinionated, and directly usable in day-to-day sprint work.
1. STORY READING & ESTIMATION — Explain step-by-step how every developer must read, understand, and estimate user stories during the weekly Backlog Grooming session (before stories are allocated to a sprint). Include: a 6-step grooming process, the Fibonacci Planning Poker card reference with descriptions, the 5W clarification model, and techniques for splitting large stories.
2. DEFINITION OF READY (DoR) — Write an 8-point DoR checklist that defines precisely when a story is ready to enter a sprint. Cover: story format (INVEST), acceptance criteria (Given-When-Then), estimation, UX wireframe agreement, technical approach agreement, test approach, dependency resolution, and zero open blockers. Include a Given-When-Then writing guide with real flight booking examples.
3. DEFINITION OF DONE (DoD) — Write a comprehensive DoD guide covering 4 stages:
– Design DoD: How UX design must be completed (wireframes, accessibility, handoff)
– Coding DoD: How code must be written (ACs, standards, no hardcodes, logging)
– Code Review DoD: How peer review must be completed (PR standards, two-reviewer rule, security check)
– Testing DoD: Unit / Integration / System / Regression / UAT — with pass criteria and artefacts for each.
End with a 7-point Final DoD Gate checklist.
4. DEVELOPMENT WORKFLOW — Show a 12-stage software development workflow from “Story pulled into sprint” to “Merged and demo-ready.” For each stage: describe developer actions, the DoD gate that must pass before proceeding, who is responsible, and the sprint day timing. Include a developer daily time plan.
5. SPRINT DEMO GUIDE — Explain with detailed worked examples how developers present sprint demos so that:
– The Product Owner can validate User Requirement stories against acceptance criteria
– The Tech Lead can validate Enabler Requirement stories against technical criteria Provide 2 fully scripted demo examples (one user story, one enabler story) showing exact developer dialogue, what the validator looks for, and how to handle accepted / rejected / partially accepted outcomes.
// Format: A professional Word document with colour-coded stage tables, info boxes,
// checklist tables, and a cover page. Practical tone — write as a developer would speak.
Why This Prompt Works
This prompt uses persona precision — “senior developer” rather than just “developer” — which activates the AI’s knowledge of code quality standards, review processes, and team mentorship. The explicit staging of DoD into 4 distinct phases (Design → Code → Review → Testing) prevents the common mistake of producing a single undifferentiated checklist. The “fully scripted demo examples” instruction is the key differentiator — it forces the AI to produce realistic, usable dialogue rather than abstract guidance. Specifying both “Product Owner validates user stories” and “Tech Lead validates enabler stories” teaches the AI the distinction between the two validation paths.
Key Outputs This Prompt Generates
6-step grooming process
8-point DoR checklist
4-stage DoD framework
12-stage dev workflow
2 scripted demo examples
Tips to Personalise This Prompt
Add your tech stack
Add “Our tech stack is: React frontend, Node.js backend, PostgreSQL, deployed on AWS using GitHub Actions CI/CD.” The DoD and workflow sections will reference your actual tools.
Change velocity expectations
Specify “our team completes an average of 5 stories per sprint” to calibrate the sprint day timing in the workflow to your team’s actual pace.
Specify test frameworks
Add “We use Jest for unit tests, Cypress for E2E tests, and Postman for API testing.” The testing DoD will reference your actual test tooling.
Request junior-focused version
Add “Write for a junior developer audience — include examples for every concept and avoid assuming Scrum experience.” Ideal for onboarding new team members.
How to Use These Prompts in 5 Steps
Open Your AI Tool
Open Claude (claude.ai) or ChatGPT (chat.openai.com) in your browser. No account needed for basic access. Claude is recommended for longer document generation.
Copy the Prompt
Click the Copy button on any of the three prompts above. Paste it into the AI chat window. Personalise the project name, team size, and year before sending.
Review the Output
Read the AI's response. If any section is incomplete, type a follow-up: "Please expand Section 3 with more detail on testing types." The AI will continue refining.
Request the Document
Once satisfied, type: "Now format this as a professional Word document." Claude will generate the full .docx output with tables, colours, and formatting.
Share With Your Team
Download the document. Share with SM, PO, and developers in your next Sprint Planning or team kickoff. Update it quarterly as your team evolves.
Download the 3 Agile Guides — Free
8 Prompt Engineering Tips for Agile Teams
Always Assign a Role First
Start every prompt with "You are the [Scrum Master / Product Owner / Senior Developer]..." Role assignment is the single highest-impact technique. It shifts the AI from generic assistant to domain expert.
Number Every Sub-Task
Numbered lists in prompts produce numbered sections in outputs. "Cover: 1. Team structure 2. Sprint events 3. Timeline" guarantees complete, structured coverage — no important sections accidentally skipped.
Be Specific About Quantities
"Exactly 10 team members", "minimum 8 Epics", "24 sprints", "8-point checklist" — specificity prevents the AI from producing token examples. Vague prompts produce vague outputs.
Name Your Project
Replace generic descriptions with your actual project: "AI-enabled flight booking application" generates flight-specific examples. "E-commerce returns portal" generates retail-specific examples. Context = relevance.
Specify the Output Format
"Format as a Word document with tables, colour-coded sections, and section banners" controls presentation. Without format instructions, AI defaults to plain text — usable but not team-ready.
Use Follow-Up Prompts
After the first output, refine with: "Expand Section 3 with worked examples." "Add a column for 'Who is accountable' to the metrics table." Iteration produces the best results — don't expect perfection on the first pass.
State the Audience
"Write for a junior developer audience" vs "Write for a PMO Head presenting to the board" changes the language, depth, and structure of the output significantly. Audience matters as much as content.
Ask for Examples
Add "Include a worked example for every technique using our flight booking project" to transform abstract theory into practical, team-specific guidance. Examples are what make guides actually used.
Frequently Asked Question
You are a senior developer in a Scrum team building an AI-enabled flight booking application.
1. STORY READING & ESTIMATION
2. DEFINITION OF READY
3. DEFINITION OF DONE