smithery/kcc989

rwsdk-cron-triggers

Schedule recurring background tasks in rwsdk/Cloudflare Workers - cron syntax, setup, local testing, and patterns for cleanup, metrics, billing, and maintenance

Installation

$ npx skills add smithery/kcc989 --skill rwsdk-cron-triggers

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 5,476 B
  • docs SUMMARY.md 187 B

History

  1. First recorded snapshot · 0 installs

SKILL.md

rwsdk Cron Triggers

Schedule recurring background tasks using standard cron syntax. Tasks run automatically in production. Use for cleanup, metrics, billing, backups, and maintenance.

Quick Setup

1. Configure wrangler.jsonc

{
  "triggers": {
    "crons": [
      "0 * * * *", // Every hour
      "0 21 * * *" // Daily at 9 PM UTC
    ]
  }
}

Run pnpm generate after changes.

2. Add Scheduled Handler

export default {
  fetch: app.fetch,
  async scheduled(controller: ScheduledController) {
    try {
      switch (controller.cron) {
        case '0 * * * *': {
          await cleanupExpiredSessions();
          console.log('✅ Hourly cleanup complete');
          break;
        }
        case '0 21 * * *': {
          await processBilling();
          console.log('✅ Daily billing complete');
          break;
        }
        default: {
          console.warn(`Unhandled cron: ${controller.cron}`);
        }
      }
    } catch (error) {
      console.error('Cron error:', error);
    }
  },
} satisfies ExportedHandler<Env>;

Each case must match a cron in wrangler.jsonc.

Cron Syntax

┌─ minute (0-59)
│ ┌─ hour (0-23)
│ │ ┌─ day (1-31)
│ │ │ ┌─ month (1-12)
│ │ │ │ ┌─ weekday (0-6, Sun-Sat)
* * * * *

Common patterns:

  • * - Every minute (testing only)
  • /5 * - Every 5 minutes
  • 0 - Every hour
  • 0 0 * - Daily at midnight UTC
  • 0 9 1-5 - Weekdays at 9 AM UTC
  • 0 0 0 - Weekly (Sundays)
  • 0 0 1 - Monthly (1st day)

Operators:

  • * - Every value
  • */n - Every n units
  • , - Multiple values: 0 9,17 * (9 AM and 5 PM)
  • - - Range: 0 9-17 * (9 AM through 5 PM)

Testing Locally

Crons only fire automatically in production. Test locally with curl.

Basic test (find PORT from pnpm dev output):

curl "http://localhost:PORT/cdn-cgi/handler/scheduled?cron=0+*+*+*+*"

With custom timestamp (for time-dependent logic):

# Test as if running Jan 1, 2025 midnight UTC
curl "http://localhost:PORT/cdn-cgi/handler/scheduled?cron=0+0+1+*+*&time=1735689600"

Generate timestamps:

date +%s  # Current time
date -d "2025-01-15 12:00:00 UTC" +%s  # Specific date

Test script:

#!/bin/bash
PORT=5173  # Your dev server port

curl "http://localhost:$PORT/cdn-cgi/handler/scheduled?cron=0+*+*+*+*"
curl "http://localhost:$PORT/cdn-cgi/handler/scheduled?cron=0+21+*+*+*"

Example Patterns

Session cleanup (hourly):

"crons": ["0 * * * *"]

case "0 * * * *": {
  const result = await db
    .deleteFrom("sessions")
    .where("expiresAt", "<", new Date().toISOString())
    .executeTakeFirst();
  console.log(`🧹 Cleaned ${result.numDeletedRows} sessions`);
  break;
}

Billing (daily at 2 AM):

"crons": ["0 2 * * *"]

case "0 2 * * *": {
  const subs = await db
    .selectFrom("subscriptions")
    .where("nextBillingDate", "<=", today())
    .selectAll()
    .execute();

  // Queue work instead of processing inline
  await env.BILLING_QUEUE.sendBatch(
    subs.map(s => ({ body: { subscriptionId: s.id } }))
  );
  console.log(`💰 Queued ${subs.length} billing tasks`);
  break;
}

Cleanup old files (monthly):

"crons": ["0 3 1 * *"]  // 1st of month, 3 AM UTC

case "0 3 1 * *": {
  const cutoff = new Date();
  cutoff.setDate(cutoff.getDate() - 30);

  const { objects } = await env.R2_BUCKET.list({ prefix: "temp/" });

  for (const obj of objects) {
    if (obj.uploaded < cutoff) {
      await env.R2_BUCKET.delete(obj.key);
    }
  }
  console.log(`🗑️ Cleaned up old files`);
  break;
}

Best Practices

1. Queue heavy work

// ❌ BAD - might timeout
for (const user of await getAllUsers()) {
  await sendEmail(user);
}

// ✅ GOOD - queue work
const users = await getAllUsers();
await env.EMAIL_QUEUE.sendBatch(users.map((u) => ({ body: { userId: u.id } })));

2. Add error handling

try {
  await task();
} catch (error) {
  console.error('Cron failed:', error);
}

3. Document schedules

"crons": [
  "*/5 * * * *",  // Health checks
  "0 * * * *",    // Session cleanup
  "0 2 * * *"     // Billing
]

4. All times are UTC

'0 17 * * *'; // 5 PM UTC = 12 PM EST = 9 AM PST

Common Mistakes

  • ❌ Cron not in wrangler.jsonc
  • ❌ Forgetting pnpm generate
  • ❌ Long-running work in handler (use queues)
  • ❌ Assuming non-UTC timezone
  • ❌ No error handling or logging
  • ❌ Missing default case in switch

Key Limits

  • Minimum interval: 1 minute
  • Maximum schedules: 3 (free tier)
  • Timezone: UTC only
  • Execution: At-least-once (make idempotent)
  • Timeout: Same as Worker CPU limits

Workflow

  1. Add cron to wrangler.jsonc
  2. Run pnpm generate
  3. Add case to scheduled handler
  4. Test with curl: curl "http://localhost:PORT/cdn-cgi/handler/scheduled?cron=0++++"
  5. Deploy and monitor in Cloudflare Dashboard

Architecture

Cron Trigger → Check what needs doing → Queue work → Process in queue consumer

Keep cron handlers lightweight. Use them to schedule work, not process it.