メニュー
Simplify programming long running code flows. Workflows execute durably with configurable retries and delays. Use this skill whenever working with Workflow or related Convex component functionality.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
General development best practices and common gotchas when working on Biome. Use for avoiding common mistakes, understanding Biome-specific patterns (AST, syntax nodes, string extraction, embedded languages), and learning technical tips.
Guide for creating and writing proper changesets for Biome PRs. Use when a PR introduces user-visible changes (bug fixes, new features, rule changes, formatter changes, parser changes) that need a changeset entry for the CHANGELOG. Trigger when creating changesets, writing changeset descriptions, or choosing the correct change type.
Building AI agents with the Convex Agent component including thread management, tool integration, streaming responses, RAG patterns, and workflow orchestration
Guidelines for building production-ready Convex apps covering function organization, query patterns, validation, TypeScript usage, error handling, and the Zen of Convex design philosophy
How to create, structure, and publish self-contained Convex components with proper isolation, exports, and dependency management
Scheduled function patterns for background tasks including interval scheduling, cron expressions, job monitoring, retry strategies, and best practices for long-running tasks
| name | convex-dev-workflow |
| description | Simplify programming long running code flows. Workflows execute durably with configurable retries and delays. Use this skill whenever working with Workflow or related Convex component functionality. |
The Workflow component adds durable execution to Convex functions, allowing you to build reliable long-running processes that survive server restarts and handle failures gracefully. It combines queries, mutations, and actions into deterministic workflows with configurable retry policies, parallel execution, and reactive status monitoring. Workflows can run for months with custom delays, be canceled mid-execution, and provide real-time progress updates to multiple subscribers.
npm install @convex-dev/workflow
• User onboarding flows that span multiple days with email verification, content generation, and follow-up messages that must complete reliably • Data processing pipelines that orchestrate multiple API calls, transformations, and database updates with automatic retries on transient failures • Payment and fulfillment workflows that coordinate between payment processors, inventory systems, and shipping APIs with precise error handling • Content moderation systems that combine AI analysis, human review queues, and notification steps with configurable delays between stages • Batch job orchestration that processes large datasets by coordinating multiple parallel workers with backpressure and failure recovery
You define workflows using workflow.define() with deterministic handler functions that orchestrate Convex queries, mutations, and actions through a step context. The component restricts access to non-deterministic APIs like fetch and patches others like Math.random() to ensure reproducible execution across server restarts.
Workflows are started from mutations or actions using workflow.start(), which returns a WorkflowId for tracking progress. Steps execute via ctx.runQuery(), ctx.runMutation(), and ctx.runAction() methods, with results passed between steps. Parallel execution happens through Promise.all() calls, while retry behavior is configured per-step or globally through workpool options.
The component provides reactive status monitoring through workflow.status(), cancellation with workflow.cancel(), and completion handling via onComplete callbacks. Failed workflows can be restarted from specific steps using workflow.restart(), and the system automatically manages step parallelism and resource cleanup.