QCSD Refinement Swarm v1.0

Shift-left quality engineering swarm for Sprint Refinement sessions.

Overview

The Refinement Swarm takes user stories that passed Ideation and prepares them for Sprint commitment using SFDIPOT product factors, BDD scenarios, and INVEST validation. It renders a READY / CONDITIONAL / NOT-READY decision.

QCSD Phase Positioning

Phase Swarm Decision When

Ideation qcsd-ideation-swarm GO / CONDITIONAL / NO-GO PI/Sprint Planning

Refinement qcsd-refinement-swarm READY / CONDITIONAL / NOT-READY Sprint Refinement

Development qcsd-development-swarm SHIP / CONDITIONAL / HOLD During Sprint

Verification qcsd-cicd-swarm RELEASE / REMEDIATE / BLOCK Pre-Release / CI-CD

Production qcsd-production-swarm HEALTHY / DEGRADED / CRITICAL Post-Release

Parameters

• STORY_CONTENT : User story with acceptance criteria (required)

• OUTPUT_FOLDER : Where to save reports (default: ${PROJECT_ROOT}/Agentic QCSD/refinement/ )

ENFORCEMENT RULES - READ FIRST

Rule Enforcement

E1 MUST spawn ALL THREE core agents in Step 2.

E2 MUST put all parallel Task calls in a SINGLE message.

E3 MUST STOP and WAIT after each batch.

E4 MUST spawn conditional agents if flags are TRUE.

E5 MUST apply READY/CONDITIONAL/NOT-READY logic exactly.

E6 MUST generate the full report structure.

E7 Each agent MUST read its reference files before analysis.

E8 MUST apply qe-test-idea-rewriter transformation in Step 8.

E9 MUST execute Step 7 learning persistence.

Step Execution Protocol

Execute steps sequentially by reading each step file with the Read tool.

Steps

• Flag Detection -- steps/01-flag-detection.md -- Analyze story content, evaluate all 7 flags

• Core Agents -- steps/02-core-agents.md -- Spawn qe-product-factors-assessor, qe-bdd-generator, qe-requirements-validator

• Batch 1 Results -- steps/03-batch1-results.md -- Wait and extract metrics

• Conditional Agents -- steps/04-conditional-agents.md -- Spawn flagged agents

• Decision Synthesis -- steps/05-decision-synthesis.md -- Apply READY/CONDITIONAL/NOT-READY logic

• Report Generation -- steps/06-report-generation.md -- Generate refinement report

• Learning Persistence -- steps/07-learning-persistence.md -- Store findings to memory

• Transformation -- steps/08-transformation.md -- Run test idea rewriter on all test ideas

• Final Output -- steps/09-final-output.md -- Display completion summary

Execution Instructions

• Use the Read tool to load the current step file

• Execute the step's instructions completely

• Verify all success criteria are met

• Pass output as context to next step

• If a step fails, halt and report

Resume Support

To resume from a specific step: specify --from-step N .

Agent Inventory

Agent Type Domain Batch

qe-product-factors-assessor Core requirements-validation 1

qe-bdd-generator Core requirements-validation 1

qe-requirements-validator Core requirements-validation 1

qe-contract-validator Conditional (HAS_API) contract-testing 2

qe-impact-analyzer Conditional (HAS_REFACTORING) code-intelligence 2

qe-dependency-mapper Conditional (HAS_DEPENDENCIES) code-intelligence 2

qe-middleware-validator Conditional (HAS_MIDDLEWARE) enterprise-integration 2

qe-odata-contract-tester Conditional (HAS_SAP_INTEGRATION) enterprise-integration 2

qe-sod-analyzer Conditional (HAS_AUTHORIZATION) enterprise-integration 2

qe-test-idea-rewriter Transformation (always) test-generation 3

Total: 10 agents (3 core + 6 conditional + 1 transformation)

Key Principle

Refinement quality determines sprint success. This swarm ensures stories are testable, complete, and ready for development commitment.