Imported from proffesor-for-testing/agentic-qe (.claude/skills/qcsd-production-swarm/SKILL.md). Install upstream with npx skills add proffesor-for-testing/agentic-qe --skill qcsd-production-swarm. Copyright stays with the author.

QCSD Production Swarm v1.0

Post-release production health assessment and QCSD feedback loop closure.

Overview

The Production Swarm assesses release health in the live production environment using DORA metrics, incident RCA, defect prediction, and cross-phase feedback loops. It renders a HEALTHY / DEGRADED / CRITICAL decision and is the only QCSD phase with dual responsibility: assessing current production health AND closing the feedback loop back to Ideation and Refinement phases.

QCSD Phase Positioning

Parameters

• TELEMETRY_DATA: Path to production telemetry, incident reports, and DORA metrics (required)
• RELEASE_ID: Release identifier for tracking (optional)
• OUTPUT_FOLDER: Where to save reports (default: ${PROJECT_ROOT}/Agentic QCSD/production/)
• SLA_DEFINITIONS: Path to SLA/SLO target definitions (optional)

ENFORCEMENT RULES - READ FIRST

PROHIBITED BEHAVIORS:

• Summarizing instead of spawning agents
• Skipping agents "for brevity"
• Proceeding before background tasks complete
• Providing your own analysis instead of spawning specialists
• Omitting report sections or using placeholder text

Step Execution Protocol

This skill uses a micro-file step architecture. Each step is a self-contained file loaded one at a time to avoid "lost in the middle" context degradation.

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

Steps

1. Flag Detection -- steps/01-flag-detection.md -- Retrieve CI/CD signals, detect telemetry source, evaluate all 7 flags
2. Core Agents -- steps/02-core-agents.md -- Spawn qe-metrics-optimizer, qe-defect-predictor, qe-root-cause-analyzer in parallel
3. Batch 1 Results -- steps/03-batch1-results.md -- Wait for core agents, extract all metrics
4. Conditional Agents -- steps/04-conditional-agents.md -- Spawn flagged conditional agents in parallel
5. Decision Synthesis -- steps/05-decision-synthesis.md -- Apply HEALTHY/DEGRADED/CRITICAL logic
6. Report Generation -- steps/06-report-generation.md -- Generate executive summary and full report
7. Learning Persistence -- steps/07-learning-persistence.md -- Store findings to memory, save persistence record
8. Feedback Loop -- steps/08-feedback-loop.md -- Run learning coordinator then transfer specialist (sequential)
9. Final Output -- steps/09-final-output.md -- Display completion summary with all scores

Execution Instructions

1. Use the Read tool to load the current step file (e.g., Read({ file_path: ".claude/skills/qcsd-production-swarm/steps/01-flag-detection.md" }))
2. Execute the step's instructions completely
3. Verify all success criteria are met before proceeding
4. Pass the step's output as context to the next step
5. If a step fails, halt and report the failure point -- do not skip ahead

Resume Support

To resume from a specific step: specify --from-step N and the orchestrator will skip to step N. Ensure you have the required prerequisite data from prior steps.

Agent Inventory

Total: 12 agents (3 core + 7 conditional + 2 feedback)

Quality Gate Thresholds

Report Filename Mapping

Execution Model Options

Key Principle

Production health is measured by outcomes, not intentions. This swarm provides evidence-based production assessment and closes the QCSD feedback loop.