OpenClaw demos impress but still no blank check: weeks of testing keep the author in Claude Code's loop
Cheapest_NOOb · reddit · 2026-10-16
After running OpenClaw alongside Claude Code for weeks, the author shares: the demos are the easy part — an agent that keeps working while you're away and messages you. But for work they'd actually stand behind, they still do what they do in Claude Code: make it plan first, read the plan, and never auto-accept runs.
Real-world notes: first reply was 2 seconds when the box was healthy, but the gateway died overnight three times, taking the cool parts down with it. The open question: is OpenClaw worth it for real work, or is the useful version still "Claude Code, with me in the loop"? They ask what others let it do without reading the plan first.
More from coding & agent
- Claude Code Pushes Urgent Security Update Across All Installations — umeshonai · 2026-10-23
- 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