Cursor Super + Claude Pro + OpenCode: can this stack survive heavy autonomous coding?
DarioAmodeiWasTaken · reddit · 2026-10-18
A heavy developer on Reddit asks whether his AI coding subscription stack makes sense for fully autonomous work (repo exploration, file edits, test runs, self-fixing) rather than autocomplete. Plan: Cursor Super as the main agent (15x Plus usage, priority access, bundled X Premium+), Claude Pro only for planning, architecture discussions, diff reviews and turning ideas into specs, and OpenCode Go Plus as a second terminal agent for batch changes or alternate models. His key concern: whether Cursor Super's weekly pool (covering only Cursor/Grok models like Grok 4.7/4.6/4.5 and Composer 2.5) survives heavy daily agent use, since Claude/GPT/Gemini inside Cursor are billed separately. He also weighs upgrading to Cursor Ultra, whether Claude Pro suffices for planning-only use, and whether OpenCode is redundant.
More from coding & agent
- Claude-written test harness let 31KB fake archives land on a real backup server — Grimmoner · 2026-10-18
- Matt Pocock's pre-commit hook caps agent-written files at 5,000 tokens — mattpocockuk · 2026-10-18
- Backups silently failed for 10 days: three monitoring blind spots lined up — Grimmoner · 2026-10-18
- Author Laurence Moroney: write your ugly first draft yourself, then let an LLM edit it — lmoroney · 2026-10-18
- Open-weight TYGER speech decision model makes 80ms turn-taking calls for full-duplex voice agents — solyarisoftware · 2026-10-18
- Devs: Opus 5.5 on medium beats Opus on max — eyishazyer · 2026-10-18