Gathering Requirements

Requirements Gathering

5-layer requirements elicitation for Cursor agent system generation with axioms, purpose, and methodology

Executes a structured multi-phase questionnaire implementing the 5-layer architecture to capture all requirements for generating a new Cursor agent development system.

Philosophy

This skill guides users from foundational axioms through purpose, principles, methodology, and technical implementation - ensuring generated systems are grounded in clear values and purpose.

Process

1. Review the task requirements.
2. Apply the skill's methodology.
3. Validate the output against the defined criteria.

Step 1: Axiom Configuration (Layer 0)

Select core axioms (A1-A5) and optional axioms (A6-A10) that will govern the agent system. Axiom 0 (Love, Truth, and Beauty) is always included.

Step 2: Purpose Definition (Layer 1)

Define mission statement, identify primary stakeholders, and establish measurable success criteria.

Step 3: Depth Selection

Choose Quick Start, Standard, or Comprehensive depth level. This determines which subsequent layers will be configured.

Step 4: Principles Selection (Layer 2, if Standard+)

Define ethical boundaries, quality standards, and failure handling approaches.

Step 5: Methodology Selection (Layer 3, if Standard+)

Choose development methodology (Agile Scrum, Kanban, R&D, Enterprise) and configure team size and workflow patterns.

Step 6: Enforcement & Practices (if Comprehensive)

Select enforcement patterns for quality, safety, and integrity, plus daily/craft/alignment practices.

Step 7: Technical Configuration (Layer 4)

Gather project context, technology stack, workflow triggers, knowledge domain, and agent capabilities.

Step 8: Validation & Generation

Validate completeness, confirm target directory, generate summary, and create all artifacts.

Complete Phase Structure

┌─────────────────────────────────────────────────────────────────┐
│ PRE-PHASE: Layer 0 - Axiom Configuration                        │
│   Select core (A1-A5) and optional (A6-A10) axioms              │
├─────────────────────────────────────────────────────────────────┤
│ PHASE 0: Layer 1 - Purpose Definition                           │
│   Mission, Stakeholders, Success Criteria                        │
├─────────────────────────────────────────────────────────────────┤
│ PHASE 0.5: Depth Selection                                       │
│   Quick Start / Standard / Comprehensive                         │
├─────────────────────────────────────────────────────────────────┤
│ PHASE 0.6: Layer 2 - Principles (if Standard+)                   │
│   Ethical boundaries, quality standards, failure handling        │
├─────────────────────────────────────────────────────────────────┤
│ PHASE 0.7: Layer 3 - Methodology (if Standard+)                  │
│   Agile/Kanban/R&D/Enterprise selection and configuration        │
├─────────────────────────────────────────────────────────────────┤
│ PHASE 0.8: Enforcement Selection (if Comprehensive)              │
│   Quality, Safety, Integrity enforcement patterns                │
├─────────────────────────────────────────────────────────────────┤
│ PHASE 0.9: Practice Selection (if Comprehensive)                 │
│   Daily, Craft, Alignment practices                              │
├─────────────────────────────────────────────────────────────────┤
│ PHASE 1-5: Layer 4 - Technical Configuration                     │
│   Stack, Agents, Skills, Knowledge, Integrations                 │
└─────────────────────────────────────────────────────────────────┘

PRE-PHASE: Layer 0 - Axiom Configuration

Skill Reference: {directories.skills}/axiom-selection/SKILL.md

Introduction

Before we begin, let's ground ourselves in the foundation of foundations:

Axiom 0: Love, Truth, and Beauty
"All being and doing is grounded in love, truth, and beauty."

This axiom is always included and cannot be removed. From this foundation,
we derive the core axioms that govern agent behavior.

Foundation Axiom (Always Included)

Core Axioms (Always Included)

Optional Axioms

Output

layer0:
  foundationAxiom: "A0"  # Love, Truth, and Beauty - always included
  coreAxioms: ["A1", "A2", "A3", "A4", "A5"]
  optionalAxioms: ["A6", "A7"]  # User selected
  customAxioms: []  # Domain-specific additions

PHASE 0: Layer 1 - Purpose Definition

Skill Reference: {directories.skills}/purpose-definition/SKILL.md

Questions

| # | Question | Variable | Validation | ||-|-|| | 1 | In ONE sentence, why should this agent system exist? | {MISSION_STATEMENT} | Verifiable outcome (A1) | | 2 | Who are the primary users or beneficiaries? | {PRIMARY_STAKEHOLDERS} | Specific humans (A2) | | 3 | What is the single most important outcome? | {SUCCESS_CRITERIA} | Measurable (A1) |

Output

layer1:
  mission: "{MISSION_STATEMENT}"
  stakeholders: "{PRIMARY_STAKEHOLDERS}"
  successCriteria: "{SUCCESS_CRITERIA}"

Generated Artifact: PURPOSE.md

PHASE 0.5: Depth Selection

Question

How deep should we define the remaining layers?

A) Quick Start - Use defaults, go straight to technical configuration
B) Standard - Define principles and select methodology template
C) Comprehensive - Define all layers including enforcement and practices

Routing

| Selection | Next Phase | |--|| | Quick Start | Phase 1 (Technical) | | Standard | Phase 0.6 (Principles) | | Comprehensive | Phase 0.6 (Principles) |

PHASE 0.6: Layer 2 - Principles (Standard & Comprehensive)

Questions

| # | Question | Variable | Reference | ||-|-|--| | 1 | What actions should agents NEVER take? | {ETHICAL_BOUNDARIES} | {directories.patterns}/principles/ethical-boundaries.json | | 2 | What quality bar must all outputs meet? | {QUALITY_STANDARDS} | {directories.patterns}/principles/quality-standards.json | | 3 | How should failures be handled? | {FAILURE_HANDLING} | {directories.patterns}/principles/failure-handling.json |

Output

layer2:
  ethicalBoundaries: ["EB1", "EB2", "EB3", "EB4", "EB5"]
  qualityStandards: ["QS1", "QS2", "QS3", "QS4", "QS5"]
  failureHandling: ["FH1", "FH2", "FH3"]

PHASE 0.7: Layer 3 - Methodology (Standard & Comprehensive)

Skill Reference: {directories.skills}/methodology-selection/SKILL.md

Questions

| # | Question | Variable | Options | ||-|-|| | 1 | What development methodology fits your team? | {METHODOLOGY} | Agile Scrum, Kanban, R&D, Enterprise | | 2 | Team size? | {TEAM_SIZE} | 1-3, 4-6, 7-10, 10+ |

Methodology Options

| Methodology | Best For | Pattern | |-|-|| | Agile Scrum | Product development, feature teams | domain_expert_swarm + peer_collaboration | | Kanban | Support, maintenance, ops | continuous_flow + pull_based | | R&D | AI/ML, innovation, research | knowledge_mesh + adaptive_learning | | Enterprise | Large-scale, compliance-driven | hierarchical_command + pipeline |

Output

layer3:
  methodology: "agile-scrum"
  teamSize: "4-6"
  pattern: "domain_expert_swarm + peer_collaboration"

Generated Artifact: {directories.workflows}/methodology.yaml

PHASE 0.8: Enforcement Selection (Comprehensive Only)

Skill Reference: {directories.skills}/enforcement-selection/SKILL.md

Question

Select enforcement patterns to ensure values are lived:

Quality Enforcement:
[x] E1: Test Coverage Gate
[x] E2: Peer Review Gate
[ ] E3: Documentation Completeness
[x] E4: Style Consistency

Safety Enforcement:
[x] E5: Destructive Action Confirmation
[ ] E6: Backup Before Modification
[ ] E7: Security Vulnerability Check
[ ] E8: Production Safeguard

Integrity Enforcement:
[x] E9: Axiom Compliance Check
[ ] E10: Purpose Alignment Check
[ ] E11: Transparency Log

Output

enforcement:
  quality: ["E1", "E2", "E4"]
  safety: ["E5"]
  integrity: ["E9"]

Generated Artifact: enforcement.yaml

PHASE 0.9: Practice Selection (Comprehensive Only)

Skill Reference: {directories.skills}/practice-selection/SKILL.md

Question

Select practices to maintain excellence and alignment:

Daily Practices:
[ ] P1: Morning Intention Setting
[ ] P2: Evening Reflection
[x] P3: Focused Stand-up

Craft Practices:
[x] P4: Code as Craft Review
[x] P5: Thoughtful Code Review
[ ] P6: Continuous Refactoring

Alignment Practices:
[ ] P7: Weekly Learning Session
[x] P8: Sprint Retrospective
[ ] P9: Release Blessing
[ ] P10: Quarterly Purpose Alignment

Output

practices:
  daily: ["P3"]
  craft: ["P4", "P5"]
  alignment: ["P8"]

Generated Artifact: practices.yaml

PHASE 1: Project Context (Layer 4 - Technical)

Questions

| # | Question | Variable | Validation | ||-|-|| | 1 | What is the name of your project? | {PROJECT_NAME} | Valid directory name | | 2 | Briefly describe what this project will do | {PROJECT_DESCRIPTION} | Non-empty, < 500 chars | | 3 | What domain or industry is this for? | {DOMAIN} | Match to known domains | | 4 | Team size and experience level? | {TEAM_CONTEXT} | Optional |

Domain Options

• Web Development
• Mobile Development
• Data Science / ML
• AI Agent Development
• SAP / Enterprise
• DevOps / Infrastructure
• Game Development
• IoT / Embedded
• Custom (specify)

Output

projectContext:
  name: "{PROJECT_NAME}"
  description: "{PROJECT_DESCRIPTION}"
  domain: "{DOMAIN}"
  teamContext: "{TEAM_CONTEXT}"

PHASE 2: Technology Stack

Questions

| # | Question | Variable | Reference | ||-|-|--| | 1 | Primary programming language? | {PRIMARY_LANGUAGE} | Check blueprints | | 2 | Frameworks or libraries? | {FRAMEWORKS} | Stack-specific | | 3 | Database or storage systems? | {DATABASES} | Optional | | 4 | External APIs or services? | {EXTERNAL_APIS} | Optional |

Blueprint Matching

IF primaryLanguage == "python" AND "fastapi" IN frameworks:
    suggest blueprint: "python-fastapi"
IF primaryLanguage == "python" AND ("langchain" IN frameworks OR "langgraph" IN frameworks):
    suggest blueprint: "ai-agent-development"
IF primaryLanguage == "typescript" AND "react" IN frameworks:
    suggest blueprint: "typescript-react"

Output

technologyStack:
  primaryLanguage: "{PRIMARY_LANGUAGE}"
  frameworks: ["{FRAMEWORK_1}", "{FRAMEWORK_2}"]
  databases: ["{DATABASE_1}"]
  externalApis: ["{API_1}"]
  matchedBlueprint: "{BLUEPRINT_ID}" | null

PHASE 3: Workflow Methodology

Questions

| # | Question | Variable | Options | ||-|-|| | 1 | What triggers tasks? | {TRIGGER_SOURCES} | Jira, Confluence, GitHub, GitLab, Manual | | 2 | Output artifacts? | {OUTPUT_ARTIFACTS} | Code, Docs, Tests, Diagrams |

Trigger Source Details

| Trigger | Additional Info Needed | ||-| | Jira | Project key pattern (e.g., PROJ-###) | | Confluence | Space key | | GitHub | Repository URL pattern | | GitLab | Repository URL pattern |

Output

workflowConfig:
  triggerSources:
    - type: "jira"
      pattern: "{PROJECT_KEY}-{NUMBER}"
  outputArtifacts: ["code", "docs", "tests"]

PHASE 4: Knowledge Domain

Questions

| # | Question | Variable | Purpose | ||-|-|| | 1 | Domain-specific concepts? | {DOMAIN_CONCEPTS} | Build knowledge files | | 2 | Reference repositories/docs? | {REFERENCE_SOURCES} | DeepWiki integration | | 3 | Naming conventions? | {CONVENTIONS} | Style guide setup |

Output

knowledgeDomain:
  concepts: ["{CONCEPT_1}", "{CONCEPT_2}"]
  referenceSources:
    - type: "github"
      url: "{REPO_URL}"
  conventions:
    styleGuide: "{STYLE_GUIDE}"

PHASE 5: Agent Capabilities

Questions

| # | Question | Variable | Options | ||-|-|| | 1 | Core agents needed? | {CORE_AGENTS} | Multi-select | | 2 | Workflow skills needed? | {CORE_SKILLS} | Multi-select | | 3 | MCP server integrations? | {MCP_INTEGRATIONS} | Multi-select |

Agent Options

Skill Options

Output

agentCapabilities:
  agents: ["code-reviewer", "test-generator"]
  skills: ["bugfix-workflow", "feature-workflow", "tdd"]
  mcpIntegrations: ["atlassian", "deepwiki"]

Final Validation & Summary

Completeness Check

• Layer 0: Axioms configured
• Layer 1: Purpose defined
• Layer 2: Principles selected (if Standard+)
• Layer 3: Methodology chosen (if Standard+)
• Layer 4: Technical stack defined
• At least one agent selected
• At least one skill selected

Target Directory

1. Ask: "Where should I create the project?"
2. Validate: Path is accessible, NOT within factory directory
3. Confirm with user

Generate Summary

## Requirements Summary

### Layer 0: Axioms
- **Core:** A1-A5 (Verifiability, User Primacy, Transparency, Non-Harm, Consistency)
- **Optional:** {SELECTED_OPTIONAL_AXIOMS}

### Layer 1: Purpose
- **Mission:** {MISSION_STATEMENT}
- **Stakeholders:** {PRIMARY_STAKEHOLDERS}
- **Success:** {SUCCESS_CRITERIA}

### Layer 2: Principles
- **Boundaries:** {ETHICAL_BOUNDARIES}
- **Standards:** {QUALITY_STANDARDS}

### Layer 3: Methodology
- **Methodology:** {METHODOLOGY}
- **Team Size:** {TEAM_SIZE}

### Layer 4: Technical
- **Project:** {PROJECT_NAME}
- **Stack:** {PRIMARY_LANGUAGE} + {FRAMEWORKS}
- **Blueprint:** {MATCHED_BLUEPRINT}
- **Agents:** {CORE_AGENTS}
- **Skills:** {CORE_SKILLS}

### Output Location
- **Target:** {TARGET_DIR}

Generated Artifacts

Team Alternative

For teams of 2+ people, offer the Team Workshop Onboarding:

"Would you prefer to gather requirements:

A) Individually - I'll ask you questions now (faster, solo decision-maker)
B) As a Team - Workshop series with games and discussions (richer, collaborative)

Team workshops are grounded in Axiom 0 (Love, Truth, and Beauty) and include:
- Vision Quest: Headlines and stakeholder games
- Ethics Arena: Dilemma debates and value auctions
- Stack Safari: Trade-off games and architecture pictionary
- Agent Assembly: Trading cards and skill bingo
- Integration Celebration: Demo derby and gratitude circle"

If they choose team workshops, hand off to team-workshop-onboarding skill.

Important Rules

1. Ask one phase at a time - Don't overwhelm user
2. Explain the why - Connect questions to axioms and purpose
3. Validate inputs - Check against axioms before proceeding
4. Suggest blueprints - Match to existing blueprints
5. Confirm summary - Review before generation
6. Never assume - Ask if uncertain (A2: User Primacy)
7. Ground in A0 - Remember that love, truth, and beauty underpin all requirements

References

• {directories.patterns}/axioms/ - Axiom definitions
• {directories.patterns}/principles/ - Principle patterns
• {directories.patterns}/methodologies/ - Methodology templates
• {directories.patterns}/enforcement/ - Enforcement patterns
• {directories.patterns}/practices/ - Practice patterns
• {directories.knowledge}/ - Domain knowledge files
• {directories.blueprints}/ - Technology blueprints

When to Use

This skill should be used when strict adherence to the defined process is required.

Prerequisites

• Basic understanding of the agent factory context.
• Access to the necessary tools and resources.

Best Practices

• Always follow the established guidelines.
• Document any deviations or exceptions.
• Regularly review and update the skill documentation.