Don't Automate Everything: Help Your Coding and Research Agents Get Better First
omarsar0 · x · 2026-11-21
The author reflects on an early mistake in software engineering and research: trying to automate everything at once with a pile of arbitrary, subjective skills. Now he spends most of his time helping his coding agents write better code and his research agents explore novel research ideas — meaning he is more involved, not less. He offered to put together a list of resources and tools to help others apply this approach in their own workflows.
Related event: Focus on Helping Agents Write Better Code Before Full Automation(2 posts)→
More from coding & agent
- Animama: open-source AI desktop companion with VRM avatars and versioned memory files — kavidy · 2026-11-22
- Building an AI team with Claude: an AI employee that already knows its job — blaizedsouza · 2026-11-22
- SteerCamp ships Mac app: delegate work to a crew of AI members on cloud machines — pramodk73 · 2026-11-22
- Open-source Claude Code plugin cuts complete music videos from a single song, now at v1.7 — AIandDesign · 2026-11-22
- Watch a full AAA Batman Glider game being built from scratch in Spawn — majidmanzarpour · 2026-11-22
- Karpathy's Coding Rules Packed Into One CLAUDE.md Hit 232K GitHub Stars — blaizedsouza · 2026-11-22