Core responsibilities
- Gather and refine requirements from stakeholders
- Write and own user stories and acceptance criteria in Jira
- Drive solution design at the individual story level
- Contribute to delivery through configuration (click, not code)
- Support sprint planning and backlog grooming with the PO
- Create and maintain documentation for QA and Support
Where AI saves you the most time
- Generating user stories from requirements docs (hours → minutes)
- Creating user stories directly from Figma designs via MCP
- Writing test scripts and test data from business requirements
- Drafting Confluence documentation from sprint demos
- Querying and creating Jira tickets in natural language
- Performing impact assessments across stories and epics
Key AI tools for this role
- GitHub Copilot + prompt files — primary daily driver
- Figma MCP — generate stories from UI designs
- Jira MCP — manage tickets via natural language
- Rovo (Jira AI) — sprint health and ticket analysis
- Figma Make — rapid prototype creation
AI across the delivery lifecycle
- Pre-sprint: Generate stories from requirements or Figma
- Sprint kickoff: Validate stories meet definition of ready
- During sprint: Jira MCP for ticket updates and tracking
- Demo: Draft demo scripts and release notes
- Post-sprint: Auto-generate Confluence documentation
.prompt.md files inside .github/prompts/ and trigger them by typing the filename in Copilot Chat. Click any template to expand it.
generate-user-stories.prompt.md
# generate-user-stories.prompt.md --- title: User Story Generation description: Generates complete stories with acceptance criteria and sizing from a requirements doc. --- # Role You are a senior Business Systems Analyst on a Salesforce CRM delivery team. Your job is to convert the provided requirements document into sprint-ready user stories. # Story Format — use this structure for every story Title: As a [persona], I want [action] so that [business value] Type: Functional | Technical | Configuration Acceptance Criteria: Given [context], When [action], Then [expected result] Story Size: S (1–3pts) | M (3–5pts) | L (8pts) | XL (13pts+, split recommended) Dependencies: [list any blockers or prerequisites] Risks / Assumptions: [list any] # Rules Follow INVEST principles. Separate functional from technical stories. Do not combine multiple features into one story. Flag any gaps or ambiguities in the requirements as open questions.
figma-to-stories.prompt.md
# figma-to-stories.prompt.md
---
title: Figma Design → User Stories
description: Analyzes a Figma design and generates user stories per screen and interaction.
---
# Role
You are a senior BSA analyzing a Figma design to extract user requirements.
# Task
For the Figma design provided:
1. Identify each distinct screen or panel
2. For each screen, identify all user interactions (clicks, inputs, navigation, validations)
3. Generate one user story per key interaction
# Output Format — per screen
Screen: [Screen name from design]
Stories:
- As a [user type], I want to [action] so that [value]
Acceptance Criteria: Given / When / Then
# Notes
Infer user type from context (Sales Rep, Manager, Admin, etc.).
Flag any interactions that appear ambiguous or incomplete in the design.
Group stories by screen for clarity.
generate-confluence-doc.prompt.md
# generate-confluence-doc.prompt.md --- title: Confluence Documentation Generator description: Turns sprint demo notes and Jira stories into structured Confluence pages. --- # Role You are a technical writer documenting a delivered feature for a Salesforce CRM platform. # Task Using the sprint notes and user stories provided, generate a Confluence documentation page. # Output Structure ## Feature Overview Brief description of what was built, and the business problem it solves. ## Functional Behavior Step-by-step description of how the feature works for each persona. ## Configuration Details Salesforce configuration changes (objects, fields, flows, permissions). ## Edge Cases & Known Limitations Behaviors to be aware of outside the happy path. ## Related Jira Stories List story IDs and titles for traceability. # Notes Write for QA engineers, Support, and future BSAs. Use plain language. Flag sections where information was insufficient.
GitHub Copilot
Code & Content AI
Your primary daily driver for user story generation, documentation, test data creation, and all prompt file workflows. Works directly inside VSCode.
When to use
Any content generation, analysis, or documentation task — every day.
Figma MCP
Design Integration
Pull live Figma designs directly into VSCode and have Copilot analyze every user interaction to generate functional user stories automatically.
When to use
When you receive a new Figma design from UX or the TPM.
Figma Make
Rapid Prototyping
Generate interactive design prototypes from a plain-English prompt in seconds. Ideal for stakeholder validation before writing stories or committing engineering effort.
When to use
Early discovery — to validate a concept before sprint commitment.
Jira MCP
Ticket Management
Create, update, query, and close Jira tickets using natural language inside Copilot Chat — manage your entire backlog without leaving VSCode.
When to use
During sprint — for ticket management and status tracking.
Rovo (Jira AI)
Sprint Intelligence
AI-powered sprint health analysis, ticket pattern recognition, and backlog insights built directly into the Jira interface.
When to use
Sprint planning, backlog grooming, and retrospective prep.
Getting set up — 6 steps
Request GitHub Copilot Access
Submit an access request through the Cisco AppStore. Install VSCode 1.109+ from the AppStore as well.
Install Copilot Extensions
In VSCode marketplace, install GitHub Copilot and GitHub Copilot Chat. Authenticate with your GitHub account.
Create Your Prompt File Structure
In your workspace root, create .github/prompts/. Add your first prompt file using the templates on the Prompt Templates tab.
Set Up Figma MCP
Request Figma developer access via AppStore. In VSCode: CTRL+SHIFT+P → "mcp" → add Figma MCP config to mcp.json. Start server and authenticate.
Configure Jira MCP
Add your Jira MCP config to mcp.json. Enable MCP tools in the Copilot Chat settings icon. Test with "list my assigned stories."
Run Your First Prompt
Drag a requirements doc into Copilot Chat, type /generate-user-stories, and review the output. Iterate in the same session — don't restart.
Prompting best practices
Assign a Role First
Start with "You are a senior BSA on a Salesforce CRM team." This dramatically improves output quality.
Drag In Your Context
Drop requirements docs or Figma PNGs directly into the Copilot Chat window for the most accurate results.
One Ask Per Prompt
Don't mix user story generation with documentation in one prompt. Run separately for better output.
Iterate, Don't Restart
Follow up in the same session: "Make the acceptance criteria more specific." Restarting loses context.
Include Examples
Add 1-2 examples of good user stories inside your prompt file. Few-shot examples are highly effective.
Always Validate Output
AI generates a strong first draft — not a final answer. Validate all stories with your PO before sprint kickoff.
Why this matters
Step-by-step guide
Tools required