AI researcher shares his workflow: agent swarms in Claude Code with harness rules
jonkhler · x · 2026-11-16
Following up on his claim that his job is nearly automatable with Claude Code, this AI researcher shares the concrete workflow: running agent swarms inside Claude Code, each equipped with a harness (i.e., a set of governing rules) and access to tools. It explains how the claimed automation of research work is actually implemented.
Related event: AI Researcher Says His Work Is Now Automated by Claude Code(2 posts)→
More from coding & agent
- Dev math: 12,000 PRs/day from an AI coding agent would cost $20K in tokens — mertdumenci · 2026-11-16
- Dev floats an MCP mapped to physical addresses so agents could query residents and businesses — mobileraj · 2026-11-16
- GitHub's new Copilot app brings repos, PRs and issues together — 0xkarasy · 2026-11-16
- Building AI agents is the easy part — developers debate what blocks revenue — SufficientLab218 · 2026-11-16
- Three practical tips for getting more out of Fable 5.2, from permission-giving to orchestrating Opus 5.5 agents — dreamwieber · 2026-11-16
- Quant researcher co-writes a short math-heavy book on agentic engineering with LLMs — blaizedsouza · 2026-11-16