Purpose: Transform AI from impressive to useful with context-aware prompts Compatible with: Claude, GPT-4, Gemini, and other AI assistants Key Principle: Always include context from previous phases
How to Use These Prompts
Pull context first: Grab the latest artifacts from 00-05 folders so the AI reflects reality.
Tell the AI who you are: Mention whether you’re acting as product, architecture, QA, or ops so tone matches.
Review like an engineer: AI output is a draft—validate feasibility, ethics, and compliance.
Commit back to context: Anything you keep must land in the context folder with traceability.
Adapt Prompts to Your Team
Startups < 50 people
Use short context snippets; reference real customers rather than formal KPIs.
Run prompts directly inside your IDE/terminal for speed.
Archive AI output in Git to build your first playbook library.
Growth-Stage / Multi-team
Embed context-folder links in every prompt so squads don’t diverge.
Rotate a “prompt steward” per sprint to review/approve shared prompts.
Pair prompts with Jira templates so work items trace to AI output.
Enterprise / Regulated
Route prompts through approved AI environments; redact PII before paste.
Add compliance checkpoints (SOX, HIPAA, etc.) inside the prompt body.
Save prompt + response as evidence inside 00/04 folders for audits.
Summarize these notes into a Phase 0 brief covering:
- core business problem & KPI target
- target users/stakeholders
- constraints + top 3 risks
Role cues: Product highlights customer signal + KPI impact; Executives confirm ROI + risk appetite; Ops/Compliance list approvals or policies that must be honored.
Sample Output (excerpt):
Problem: Claims backlog costing $2.4M/quarter.
Target KPI: Cycle time 14d → 5d by Q3.
Users: Field adjusters + compliance.
Constraints: SOC2 logging, $450k budget, legal sign-off.
Executive Intent Extraction
Extract Business Context from Documents
Use this when you have meeting notes, emails, or strategy documents and need to extract actionable business context.
Quick Start:
Summarize the attached notes into a Phase 0 brief with:
- core problem & KPI
- target users
- constraints & risks
Make it bullet-friendly for AI/to repo.
I need to extract and structure business context from the following information:
[PASTE: Meeting notes, emails, strategy documents, stakeholder conversations]
Please extract and structure this into a BVDLC Phase 0 artifact with:
1. **Core Business Problem** (2 sentences max)
- What's broken and why it matters
- Root cause, not just symptoms
2. **Target Users/Stakeholders**
- Who is affected by this problem
- Who benefits from solving it
3. **Success Metrics with Specific Targets**
- Primary KPI: Current baseline → Target value by [date]
- Secondary KPIs (if applicable)
- How we'll measure success
4. **Constraints and Non-Negotiables**
- Technical constraints
- Business constraints (budget, timeline, resources)
- Compliance/regulatory requirements
5. **Timeline and Dependencies**
- Critical dates
- External dependencies
- Market/competitive pressures
6. **Stakeholder Expectations**
- What does the executive sponsor expect?
- What are the key concerns?
7. **Risk Register**
- What could go wrong?
- How likely? How severe?
- Mitigation strategies
Format this as a structured document that AI can consume in subsequent phases.
Make it specific and measurable—no vague goals.
Phase 1: Ideation & Rapid Prototyping
Quick Start:
Draft a 4-hour prototype plan that lists:
- must-have flows
- stack suggestion
- 3-question user test script
- success criteria tied to KPI
Role cues: Product frames the value hypothesis; Engineering forces feasibility + tech stack; UX/Research ensures the user test plan is concrete.
Use this when Phase 0 is complete and you're ready to build working prototypes to validate feasibility.
**Phase 0 Business Context:**
- Problem: [BUSINESS PROBLEM FROM PHASE 0]
- Target Users: [WHO THIS IS FOR]
- Success Metrics: [PRIMARY KPI TO MOVE]
- Constraints: [KEY CONSTRAINTS]
***Task:** Build a working prototype that proves feasibility and validates value hypothesis.
Build a working prototype that:
1. Demonstrates the core value proposition
2. Uses real or realistic data (not dummy data)
3. Can be tested with actual users
4. Can be built in < 4 hours
5. Proves technical feasibility
**Technology Stack:** [SPECIFY IF KNOWN, OR ASK AI TO RECOMMEND]
**Prototype Requirements:**
- Must solve: [CORE PROBLEM]
- Must include: [MINIMUM FEATURES]
- Must prove: [KEY ASSUMPTIONS]
**Also Generate:**
1. Working prototype code
2. Setup/deployment instructions
3. User testing script with questions
4. Success criteria for validation
Make this fast and rough—we're validating direction, not building production quality.
Phase 2: Architecture & Design
Quick Start:
Convert prototype learnings into a production blueprint:
- list services + responsibilities
- note integrations + security callouts
- highlight NFR targets + deployment sketch
Role cues: Architects own trade-offs & ADRs; Security/Compliance add required controls; Product confirms architecture still delivers KPI.
Use this when architecture is complete and you need to break it into executable tasks.
**Phase 0 Business Context:**
[PASTE: Business objectives, success metrics, constraints]
**Phase 2 Architecture:**
[PASTE: System architecture, components, integration points]
Generate a comprehensive implementation plan:
**1. Component Prioritization**
Analyze which components:
- Deliver independent value (can be released alone)
- Prove Phase 0 hypothesis fastest
- Have dependencies that must be built first
- Can be built in parallel vs. sequentially
Provide prioritized build order with rationale.
**2. Task Breakdown for Each Component**
For each component, create tasks with:
| Task ID | Description | Acceptance Criteria | Input Requirements | Output Specifications | AI Suitable? | Estimated Effort | Dependencies |
|---------|-------------|--------------------|--------------------|----------------------|--------------|------------------|--------------|
| T-001 | | Functional: Business Value: | | | Yes/No | [Hours/Days] | [Task IDs] |
**Acceptance Criteria Format:**
- Functional: What it must do technically
- Business Value: How it serves Phase 0 objectives
**AI Suitability Assessment:**
- **Yes:** Clear specs, well-defined inputs/outputs, AI can generate
- **No:** Requires domain expertise, complex business logic, human judgment
**3. Execution Strategy**
- Parallel work streams: What can be built simultaneously
- Critical path: Tasks that block other work
- Risk mitigation: How we handle high-risk tasks
- Value delivery sequence: Order that delivers business value soonest
**4. AI Acceleration Plan**
For AI-suitable tasks:
- What context AI needs (link to Phase 2 docs)
- What prompts to use
- What validation humans must do
Output: Complete task backlog ready for Phase 4 implementation.
Format: Markdown tables + dependency diagram (Mermaid).
Phase 4: Build, Test & Deploy
Code Generation
Generate Implementation with Full Context
Use this when ready to implement a specific component with AI assistance.
**Phase 0 Business Context:**
[PASTE: What business problem this code solves]
**Phase 2 Architecture:**
[PASTE: Relevant architecture design, technology stack, patterns]
**Phase 3 Task:**
Task ID: [ID]
Description: [WHAT TO BUILD]
Acceptance Criteria: [FUNCTIONAL + BUSINESS VALUE REQUIREMENTS]
Input Requirements: [WHAT THIS NEEDS]
Output Specifications: [WHAT THIS PRODUCES]
**Technology Stack:** [FROM PHASE 2]
**Coding Standards:** [YOUR ORG STANDARDS OR LINK TO STYLE GUIDE]
Generate implementation for: [COMPONENT NAME]
**Requirements:**
1. **Follow Phase 2 Architecture Exactly**
- Use specified design patterns
- Adhere to component boundaries
- Match integration specifications
2. **Serve Phase 0 Business Objective**
- Every function should have a purpose traceable to business value
- No over-engineering beyond requirements
3. **Include Comprehensive Error Handling**
- Expected errors with user-friendly messages
- Unexpected errors with logging and monitoring
- Graceful degradation where appropriate
4. **Add Inline Documentation with Business Context**
- Document WHY (business reason), not just WHAT (code does)
- Include traceability comments to Phase 0/2/3
5. **Create Tests Alongside Code (TDD Approach)**
- Unit tests for business logic
- Integration tests for external dependencies
- Test coverage > 80% for value-critical paths
6. **Ensure Traceability**
- Comments linking to Phase 0 objectives
- Comments linking to Phase 2 architecture decisions
- Comments linking to Phase 3 acceptance criteria
Also generate:
- Configuration files
- Environment variables needed
- README with setup instructions and business justification
- Integration points documentation
Comprehensive Testing
Generate Test Suite
Use this when code is written and you need comprehensive test coverage.
**Phase 0 Business Objectives:**
[PASTE: Success metrics, KPIs to move]
**Phase 2 NFRs:**
[PASTE: Performance targets, security requirements, integration specs]
**Phase 4 Implementation:**
[PASTE: Brief description of what was built]
**Code Location:** [LINK OR PASTE CODE]
Generate a comprehensive test suite covering all five quality dimensions:
**1. Functional Testing**
- Happy path scenarios (expected usage)
- Edge cases from Phase 2 architecture
- Error conditions and exception handling
- User acceptance criteria from Phase 0
**2. Performance Testing**
- Load scenarios meeting Phase 2 NFR targets
- Stress testing to find breaking points
- Resource utilization (memory, CPU, network)
- Latency measurements under various loads
- Target: [SPECIFY FROM PHASE 2 NFRs]
**3. Security Testing**
- OWASP Top 10 vulnerability checks
- Authentication/authorization validation
- Input validation and sanitization
- Data protection verification (encryption, PII handling)
- Compliance requirements from Phase 0
**4. Integration Testing**
- API contracts from Phase 2
- Data flow validation across system boundaries
- External system integration points
- System boundary testing
- Mocking strategies for external dependencies
**5. Business Value Testing**
- Instrumentation to track Phase 0 KPIs
- Accuracy of business metric tracking
- Validation that code achieves Phase 0 success criteria
- A/B testing setup (if applicable)
Output:
- Executable test code with clear assertions
- Test data setup/teardown scripts
- Expected vs. actual results documentation
- Coverage report targeting >90% for value-critical paths
Format: [YOUR TEST FRAMEWORK - e.g., Jest, pytest, JUnit]
Phase 5: Monitoring & Value Tracking
Quick Start:
Set up value monitoring:
- dashboard metrics (business + tech)
- alert rules tied to KPIs
- weekly review agenda bullets
Role cues: Product/Business owns KPI narratives; Data/Analytics wire up instrumentation; Ops/SRE ensure runbooks + alerting close the loop.
Sample Output (excerpt):
Business KPI: claim cycle time → Grafana panel.
Alert: if P95 > 8h for 30 min, page duty manager.
Review agenda: KPI trend, blockers, next experiment.
Value Monitoring Setup
Set Up Value Tracking
Use this when system is deployed and you need to set up ongoing value tracking.
**Phase 0 Business Objectives:**
[PASTE: Original KPIs, success metrics, targets]
**Phase 4 Deployment:**
[PASTE: What was deployed, when, to what scale]
Set up continuous value monitoring:
**1. Business KPI Tracking**
For each Phase 0 KPI, configure:
- **Metric:** [KPI NAME]
- **Baseline:** [VALUE BEFORE DEPLOYMENT]
- **Target:** [PHASE 0 TARGET]
- **Current:** [CURRENT VALUE]
- **Trend:** [IMPROVING / FLAT / DEGRADING]
- **Measurement Method:** [HOW WE TRACK THIS]
- **Dashboard Link:** [WHERE TO VIEW]
- **Update Frequency:** [REAL-TIME / HOURLY / DAILY / WEEKLY]
**2. Value Realization Report Template**
Generate weekly report template in markdown format.
**3. Feedback Loop Triggers**
Define when to trigger feedback to earlier phases:
**Trigger Phase 0 Reassessment:**
- KPIs degrading > [THRESHOLD]% for [DURATION]
- Business assumptions proven wrong
- Market conditions changed significantly
**Trigger Phase 1 New Feature:**
- User requests for specific capability > [THRESHOLD]
- Opportunity to improve KPI by [THRESHOLD]% identified
- Competitive gap emerged
**4. Dashboard Configuration**
**Executive Value Dashboard:**
- Business KPIs with trend lines
- ROI calculation (actual vs. projected)
- User adoption metrics
- System health summary (traffic light)
**Operations Health Dashboard:**
- Technical performance metrics
- Error rates and incident history
- Resource utilization and cost
- SLA compliance
**5. Alert Rules**
**Business Metric Alerts:**
- Primary KPI degradation: [THRESHOLD LOGIC] → Notify [STAKEHOLDER]
- User adoption below target: [THRESHOLD LOGIC] → Notify [PRODUCT OWNER]
- Cost overrun: [THRESHOLD LOGIC] → Notify [FINANCE + ENGINEERING]
Output: Complete Phase 5 monitoring configuration
💡 Pro Tips for Using AI Prompts:
• Always include previous phase context—AI needs business context to be useful
• Start specific, get more specific if output is too generic
• Ask for 2-3 alternatives for AI to compare options
• Validate AI output against phase exit criteria
• Build a library of prompts that work well for your domain