Teknium runs Hermes Agent for 15 hours, sheds 375,000 lines of code
Teknium · x · 2026-09-03
Nous Research co-founder Teknium put Hermes Agent to the test with a single /goal command: the agent ran for 15 hours cleaning up the Hermes Agent repo, shedding 375,000 lines of code.
- A total of 1,320 subagents were spawned in waves of 120, plus three sets of 15 that could each spawn their own subagents.
- He calls it the ultimate stress test of Hermes Agent, demonstrating large-scale multi-agent orchestration on a real refactoring task.
Related event: Hermes Agent Runs 15 Hours, Deletes 375K Lines of Code(2 posts)→
More from coding & agent
- Non-coder runs hundreds of thousands of AI-written lines on a phone via zero-trust Termux workflow — amadale · 2026-09-03
- Multi-agent debugging pain: per-agent logs can't show which output changed the next agent's decision — mageblex · 2026-09-03
- PR adds ready-made JS and Python computer-use environments to OpenAI's CUA sample app — ChrisGPT · 2026-09-03
- 2-hour vibe-coded local agentic browser clones Opera Neon's $19/month features with MCP — comperr · 2026-09-03
- ComfyUI plugin caches MiniMax H3 conditioning to disk: 1.12s cache hits, 14GB RAM saved — Any_Fee5299 · 2026-09-03
- Open-source coding agent Pi hits 100,000 GitHub stars, v2 teased — gklambauer · 2026-09-03