Claude Opus 5 works well for coding, but it breaks Compound Engineering workflows
danshipper · x · 2026-07-25
The thread says Claude Opus 5 is useful at a medium-effort coding level, but it also breaks existing agent workflows such as Compound Engineering.
What changed in practice:
- The model works well when people stop relying on heavy skills and giant prompts.
- In an autonomous flow, it repeatedly returns control to the user, which the author finds annoying.
- The team started rewriting Compound Engineering to fit the new model behavior.
- The author also shared a custom skill for retuning their harness/plugin.
So the takeaway is not just that Opus 5 is stronger, but that agent stacks may need to be redesigned around its behavior.
Related event: Claude Opus 5 Early Tests: Better Efficiency but Overly Proactive(29 posts)→
More from coding & agent
- Is inference latency becoming the biggest bottleneck for production AI agents? — Euphoric_Sea632 · 2026-09-11
- Anthropic researcher: 99% of engineers now run swarms of 300+ self-improving agents — AlishaOutridge · 2026-09-11
- Gergely Orosz: Shipping 10x PRs With AI Agents, Sites Fill With Small Regressions — ducha_aiki · 2026-09-11
- Same Echo Maze prompt, three frontier models: all passed visually but shipped the same hidden bug — eyishazyer · 2026-09-11
- Astra storyboards plus Minimax H3 per-shot generation boost video success rates — Hailuo_AI · 2026-09-11
- Codex tip: use Sol with Astra and Luna sub-agents to save usage — pvncher · 2026-09-11