Multi-agent adversarial review experiment: Claude and Hermes negotiate a shared write lock via conversation
seanwbren · x · 2026-09-13
The author runs an experiment combining three tools: Nous Research's Hermes desktop, Anthropic's Claude desktop, and the message CLI tool ocs by @leeguooooo. The goal is collaborative, adversarial reviews between two strong models (Fable 5.1 and Astra), where a shared write lock for edits is negotiated conversationally between the agents.
Related event: Dev Runs Multi-Agent Experiment With Claude and Hermes Reviewing Each Other(2 posts)→
More from coding & agent
- 'Start more agents' is not an acceptable answer (it is): dev jokes about the idle-time problem — flavioAd · 2026-09-13
- Community ports experimental DeepSeek V4.1 steering into antirez's C-based ds4 CLI — antirez · 2026-09-13
- LangChain's guide to building a custom agent harness: context is everything — iamrobotbear · 2026-09-13
- Nex-N2.5 agentic models hit OpenRouter free tier, built around visual feedback loops — anthara_ai · 2026-09-13
- Dev to stream live demo of hooking an AI agent up to multiple MCPs — PurzBeats · 2026-09-13
- terms.txt paper proposes machine-readable access terms and pricing for AI agents — dair_ai · 2026-09-13