Compétences workflow-orchestration-patterns
🧭

workflow-orchestration-patterns

Sûr 🌐 Accès réseau⚙️ Commandes externes

Projetar fluxos de trabalho duráveis do Temporal

Transações distribuídas falham sem orquestração clara. Esta habilidade fornece padrões do Temporal para fluxos de trabalho determinísticos, retentativas e ações de compensação.

Prend en charge: Claude Codex Code(CC)
📊 71 Adéquat
1

Télécharger le ZIP du skill

2

Importer dans Claude

Allez dans Paramètres → Capacités → Skills → Importer un skill

3

Activez et commencez à utiliser

Tester

Utilisation de "workflow-orchestration-patterns". Outline a Temporal saga for booking travel with rollback steps.

Résultat attendu:

  • Register compensations before each step
  • Reserve flight, compensate by canceling reservation
  • Book hotel, compensate by canceling booking
  • Charge payment, compensate by refunding payment
  • On failure, run compensations in reverse order (LIFO)

Utilisation de "workflow-orchestration-patterns". When should I use a child workflow versus parallel activities?

Résultat attendu:

  • Use child workflows for complex, reusable process segments
  • Use parallel activities for simple independent tasks
  • Child workflows have their own execution history and timeouts
  • Activities share parent workflow history and run faster

Utilisation de "workflow-orchestration-patterns". How do I handle long-running activities that might stall?

Résultat attendu:

  • Implement heartbeat within activity to report progress
  • Set heartbeat timeout to detect stalled work
  • Use heartbeat details to enable progress-based retry
  • Break long tasks into smaller idempotent steps

Audit de sécurité

Sûr
v4 • 1/17/2026

Pure documentation skill containing only markdown guidance about Temporal workflow patterns. No executable code, scripts, or network activity. All 43 static findings are false positives triggered by documentation metadata and code block syntax - specifically git hash identifiers flagged as weak crypto, markdown backticks flagged as shell execution, and normal error classification flagged as reconnaissance.

2
Fichiers analysés
462
Lignes analysées
2
résultats
4
Total des audits

Score de qualité

38
Architecture
100
Maintenabilité
85
Contenu
30
Communauté
100
Sécurité
91
Conformité aux spécifications

Ce que vous pouvez construire

Design payment orchestration

Plan a multi-step payment workflow with compensations, retries, and idempotent activities.

Model long-lived entities

Choose entity workflows for accounts or carts with signals and queryable state.

Harden failure recovery

Define retry limits, backoff, and heartbeat policies for critical activities.

Essayez ces prompts

Workflow or activity
Classify each step of my order process as workflow or activity and explain why.
Saga checklist
Draft a saga plan for reserve inventory, charge payment, and ship order with compensations and idempotency rules.
Parallel scaling
Propose a fan-out and fan-in design for 500,000 tasks using child workflows and bounded workflow size.
Versioning strategy
Recommend a versioning approach for a long-running workflow update with backward compatibility constraints.

Bonnes pratiques

  • Keep workflows deterministic and move all I/O to activities
  • Make activities idempotent, short-lived, and always configure timeouts
  • Use child workflows to scale large parallel workloads

Éviter

  • Calling external APIs directly from workflows
  • Using system time or random functions in workflows
  • Skipping compensation registration before executing a step

Foire aux questions

Is this compatible with all Temporal SDKs?
Yes, the patterns are SDK agnostic and focus on architecture and design principles.
What are the main limits covered?
It addresses determinism, payload size, retry behavior, and long-running execution constraints.
How do I integrate this with existing services?
Use activities for external calls and keep workflows focused on orchestration and decision logic.
Does it access or store my data?
No, this is documentation only and includes no code or data handling capabilities.
What if my workflow fails repeatedly?
Use retry policies with limits, classify non-retryable errors, and add compensations for partial failures.
How does it compare to Airflow or Kafka?
It targets long-running orchestration and distributed transactions, not batch pipelines or real-time streams.

Détails du développeur

Structure de fichiers

📄 SKILL.md