Claude Code Loops Generate $100-200/Week Passive Income
Run Claude skills in a bash 'while true' loop with 'sleep 60' to automate tasks 24/7: scan Kali markets for bugs worth $25-100 each and auto-email reports, or send Hacker News digests.
Build Reusable Claude Skills for Step-by-Step Automation
Claude skills are .md files defining structured, repeatable tasks via numbered steps, invoked with claude-p /skillname in terminal for headless execution. To create one, prompt Claude Code: "fetch Anthropic Claude Code skills docs; create placeholder for skillname.md". Edit the template to outline exact steps—e.g., for bug hunting: (1) poll Kali for new prediction markets and log to new_markets.json; (2) group events by type; (3) run checklist for vulnerabilities (frontend batches, logs, exploits); (4) if bug found, email support@kali with details. Integrate Python scripts for data fetching (fetch_new_markets.py), JSON processing, and emailing via Gmail token.json. This modular setup runs reliably every cycle, avoiding ad-hoc prompts. Claude handles third-party API changes better than wrappers like OpenClaude, as /skill triggers official execution.
Infinite Bash Loop Enables 24/7 Hands-Off Operation
Wrap skill invocation in while true; do claude-p /skillname; sleep 60; done to loop indefinitely, pausing 60 seconds (or 3600 for hourly) post-run. Run in a separate terminal outside Claude Code session. Update settings.json to auto-approve bash commands: add permissions for fetch_new_markets.py, send_report.py. This yields true passivity—e.g., one loop scans Kali markets continuously, detecting minor bugs ($25 bounty), moderate ($50), severe ($100), or pre-listing extras ($10). Author nets $100-200/week from this alone; others yield $20-300 or losses (unshared). Scale by adjusting sleep for API limits, adding system prompts via claude-p --system, or chaining bash commands. GitHub repo (AllAboutAI-YT) provides open templates.
Profitable Example: Kali Bug Bounty Scanner
Target Kali's market bug bounty: poll for new prediction markets, extract logs via claim_log, analyze for exploits using fixed checklist (ensures consistency). On match, auto-email support@kali with proof. Pair watcher script logging to new_markets.json with skill consuming it—loop triggers only on fresh data, minimizing noise. Outcomes: fully autonomous, low-creativity barrier (Claude generates 80% code), runs headless. Trade-offs: rare severe bugs; on/off earnings from market volume. Replicate for any bounty/API monitoring—author runs multiples, some breakeven.
Quick Demo: Hacker News Email Digest
Prompt Claude: "Automail skill: step-by-step fetch top 5 Hacker News posts (URLs, scores), save news.json, send Gmail via token.json." Steps: (1) python fetch_hn.py for JSON; (2) send_report.py emails digest. Loop sends every 60s (demo flaw: duplicates; fix with hourly sleep). Permissions fix: settings.json allows scripts. Result: 5 emails with titles like "Built camera-only vacuum roll for <$300" link to HN. Refine to dedupe or schedule for production—proves loop scales to income tasks.