- Home
- Skills
- Bankrbot
- Claude Plugins
- Sdk Job Management
sdk-job-management_skill
- JavaScript
70
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill bankrbot/claude-plugins --skill sdk-job-management- SKILL.md2.8 KB
Overview
This skill explains how to manage asynchronous jobs in the Bankr x402 JavaScript SDK for multi-chain DeFi tasks. It covers submitting jobs, polling, checking status, cancelling, batching, and recommended timeouts so you can handle long-running or background operations reliably.
How this skill works
The SDK exposes both blocking and non-blocking methods: promptAndWait() submits and waits for completion, while prompt() returns immediately with a jobId for manual tracking. Use getJobStatus() for a single status check, pollJob() to wait until completion with a timeout, and cancelJob() to stop pending or processing jobs. Jobs follow a lifecycle: pending → processing → completed, failed, or cancelled.
When to use it
- You need to check job status or poll for results after submitting an async request.
- Requests may take longer than a single HTTP call (swaps, bridges, or complex queries).
- You want to submit work in background and retrieve results later (batch processing).
- You need ability to cancel pending or currently processing work.
- You want custom retry and timeout strategies for unstable operations.
Best practices
- Prefer promptAndWait() for simple flows where waiting is acceptable; it handles polling internally.
- Use prompt() + pollJob() for manual control, batching, or when you must offload work to background processes.
- Set timeouts based on operation type (e.g., 15s for price queries, 60–120s for cross-chain operations).
- Implement exponential backoff and a capped retry count for transient failures instead of tight tight loops.
- Always check job status before retrying or cancelling to avoid duplicate side effects.
Example use cases
- Submit a token swap and wait up to 60s for completion using promptAndWait().
- Start multiple price queries with prompt(), then poll all jobIds in parallel and aggregate results.
- Submit a balance aggregation job and periodically check status with getJobStatus() from a dashboard.
- Cancel a pending cross-chain bridge request if the user changes their mind before processing starts.
- Retry failed NFT minting requests with controlled backoff and maximum retry attempts.
FAQ
Use promptAndWait() for simple synchronous-style flows. Use pollJob() when you need manual control, advanced retry logic, or batching across many jobs.
Can I cancel a job after it starts processing?
Yes. Jobs in pending or processing states are cancellable; once completed, failed, or cancelled they cannot be changed.
How do I choose timeouts for different operations?
Follow recommended timing guidelines: short queries 15s, swaps and NFT ops 60s, cross-chain bridges 120s. Add headroom for network variability and rate limits.