atlas-iterative
Iterative 3-phase workflow with peer review cycle for changes needing validation (15-30 min)
Iterative 3-phase workflow with peer review cycle for changes needing validation (15-30 min)
Operacoes e deploy do Sr Cardoso (Cloud Run, Firestore rules/indexes, env vars, scripts, Docker). Use quando configurar ambiente, rodar deploy, revisar infra GCP ou rotinas de setup.
Validates BAZINGA completion claims with independent verification. Spawned ONLY when PM sends BAZINGA. Acts as final quality gate - verifies test failures, coverage, evidence, and criteria independently. Returns ACCEPT or REJECT verdict.
Promote spec from draft to active stage. Use after spec review passes or when ready to begin implementation.
Run an iterative architecture review loop with critique and revision
Evaluate custom stack combinations against requirements. Use when user proposes their own technology combination or asks to compare a specific stack against alternatives.
Use this skill for release preparation and execution. Invoke when deploying to staging or production environments.
Create multi-stage Azure Pipelines with deployment gates and approvals. Use when implementing complex deployment pipelines with multiple environments.
バージョン整合性チェック→提案→承認→GitHub Release作成までの標準リリースフロー(マルチエコシステム対応)
Achieve reproducible builds, dependency management, and project isolation with Nix Flakes. Use when creating reproducible environments, managing dependencies, building packages, or initializing Nix projects.
Навык запуска и сопровождения workflow bundle project-analyse (ADR-015) в системе AI-KOD: установка bundle из git/локального пути, запуск (project-analyse)/research-structure и (project-analyse)/research-analyse, требования окружения (AI_KOD_ROOT, AI_KOD_PROJECT_ROOT), контрольные точки стадий, артефакты и расследование отклонений.
Expert Deployment Engineer specializing in CI/CD automation, containerization, and release management across diverse platforms. Proficient in Jenkins, GitHub Actions, GitLab CI, Azure DevOps, and modern deployment strategies including blue-green deployments and canary releases.
Spawn 5 parallel agents applying quality principles across planning, building, and validating code. Use when designing features, making decisions, or checking work. Triggers on "launch a swarm" or when ensuring solid, maintainable, secure software.
Working in isolated worktree directories for parallel work - work normally, tests isolated, orchestrator handles cleanup
Run comprehensive audit on completed work. Use after implementation is done, or when user says 'audit this', 'review the work', 'check for issues'. Spawns specialized auditors and fixes found issues.
Execute a single phase from roadmap to working code autonomously
Pre-deployment verification for young-personal-site. Ensures build success, TypeScript, i18n sync, and functionality.
ALWAYS USE when working with Helm charts, Kubernetes deployments, kubectl commands, pod debugging, or container logs. Provides context-efficient strategies for chart development, K8s troubleshooting, and log analysis. MUST be loaded before any Helm or K8s work.
Safe Railway deployment with pre-flight guards and rollback procedures
Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always
PR 作成後に .github/workflows/ci.yml が失敗したときの対応手順。ローカルで pytest 全ケースを実行し、エラー有無で分岐。エラーがあれば修正して commit-push スキルでコミット・プッシュ、無ければ GitHub Actions のログを確認して原因を特定する。