oban-designer

What This Skill Does

The oban-designer skill is a specialized AI agent module designed to assist Elixir developers in building robust, reliable background job processing systems using the Oban library. It abstracts the complexities of worker architecture, queue management, and retry logic. Whether you are setting up your first background task or managing a high-throughput, distributed job system, this skill provides the necessary boilerplate, structural guidance, and configuration best practices. It helps you ensure that your Elixir application can handle long-running, asynchronous tasks without compromising performance or reliability.

Installation

To integrate this skill, use the ClawHub CLI: clawhub install openclaw/skills/skills/gchapim/oban. Once installed, you will need the oban dependency in your mix.exs file and the corresponding Ecto migration generated by the agent. Configure your config.exs with your repository settings and queue definitions to initialize the supervision tree properly.

Use Cases

Use this skill when your Elixir application requires: 1) Asynchronous task processing (e.g., sending emails, generating reports). 2) Handling third-party integrations via webhooks where retries are necessary. 3) Scheduled recurring maintenance tasks using Oban Pro/Cron features. 4) Resource-intensive data processing like image or video transcoding. 5) Managing complex job workflows where uniqueness constraints and priority handling are critical for system stability.

Example Prompts

1. "Create an Oban worker for processing payment receipts that retries 5 times and discards the job if the transaction is permanently rejected by the gateway."
2. "How should I configure my Oban queues to balance a high volume of webhook delivery jobs versus light internal audit log processing?"
3. "Generate a custom backoff strategy for a web scraper worker that waits longer between each attempt to avoid rate-limiting."

Tips & Limitations

• Always define unique constraints for critical jobs to prevent duplicate processing if a job is accidentally enqueued twice.
• When defining queues, align your concurrency limits with your database connection pool size to avoid saturation.
• Leverage Oban's return values effectively; distinguishing between :error (retry) and :cancel (terminate) prevents unnecessary resource drain.
• Be aware that this skill focuses on configuration and boilerplate generation. Deep architectural decisions, such as scaling Oban nodes across multiple servers, require additional infrastructure context and monitoring configuration.