I removed the approval step from my Claude Code growth agent — Reddit killed its post in under a minute
intensityflow · reddit · 2026-10-18
A solo dev shares week 4 of running a Claude Code agent that handles growth for his free Chrome extension. He removed the human approval gate: the agent now posts autonomously and only pings him on Telegram for account warnings, money, or password issues.
- He instructed it to push volume right up to the ban edge: caps raised to 5 promo posts/day, with coded tripwires in every channel that pause on the first warning sign.
- Two days in, the agent joined a subreddit and posted a gallery seconds later; Reddit's own filter removed it in under a minute. The tripwire fired, pausing all Reddit posts/replies for 72h with a Telegram ping. He lifted the pause an hour later; the account was clean, and 'don't post seconds after joining a sub' is now a rule in its playbook.
- The Product Hunt launch got 1 upvote lol; weekly users grew from 37 to 64 since the week 3 post.
- He asks others running self-posting agents: should tripwires auto-resume after a cooldown instead of waiting for a human?
More from coding & agent
- Claude cut a 2.6-hour video export to 20 minutes — but only after being told to rebuild the server — byhewar · 2026-10-23
- 3 flight-booking rules my AI agent checks that I'd never do by hand — Efistoffeles · 2026-10-23
- OpenJet claims Qwen3.8-27B coding runs 1.9x faster than llama.cpp on one RTX 3090 — JeremyCMorgan · 2026-10-23
- Building an autonomous SEO agent in Hermes with three skills: programmatic pages, Keywords Everywhere, DataForSEO — EXM7777 · 2026-10-23
- Building a sports betting AI agent with onchain execution and human approval by default: architecture and lessons — Dry-Comfortable-2678 · 2026-10-23
- How do you audit AI agent decisions in regulated workflows? A field survey — Ok_Way_1993 · 2026-10-23