cronlytic_skill

Overview

This skill provides a curl-based interface to the Cronlytic cron job scheduler API for creating, managing, pausing/resuming, and deleting scheduled HTTP requests and webhooks. It focuses on practical, reproducible curl commands and required request payloads so you can automate background tasks without running your own scheduler. Use it to schedule recurring HTTP calls, inspect logs, and control job lifecycles via simple shell commands.

How this skill works

The skill issues direct HTTP requests (curl) against Cronlytic's REST endpoints under https://api.cronlytic.com/prog/. It supports health checks, creating jobs with 5-field cron expressions, listing/updating jobs, pausing/resuming, retrieving execution logs, and deleting jobs. Authentication requires X-API-Key and X-User-ID headers set from environment variables. Example JSON payloads and jq filters are provided for common workflows.

When to use it

• Schedule recurring webhooks or API calls without managing servers
• Automate periodic background tasks like backups, health checks, or report generation
• Quickly add, pause, or resume scheduled HTTP jobs from scripts or CI pipelines
• Inspect recent job runs to debug failures or verify responses
• Manage cron expressions and job metadata programmatically via shell scripts

Best practices

• Store CRONLYTIC_API_KEY and CRONLYTIC_USER_ID in environment variables and never hard-code them
• Ping /prog/ping before heavy use to warm the API (reduces cold starts)
• Provide all required fields when creating or updating a job; Cronlytic expects complete payloads
• Validate cron expressions with a known validator to avoid scheduling mistakes
• Use jq to filter important response fields (job_id, status, next_run_at) and to inspect logs

Example use cases

• Create a POST job to trigger nightly backups at 02:00 with a single curl POST and JSON file
• Schedule a GET health-check every five minutes and pipe results into monitoring or alerting
• Pause a job automatically during deploy windows, then resume it afterward using two curl calls
• List all jobs from a maintenance script to produce an inventory of schedules and next run times
• Retrieve the last 50 execution logs for a failing job to inspect response codes and timings