Changelog Master Feed Flowing with agents (Changelog Interviews #658)
56 snips
Sep 17, 2025 
Beyang Liu, CTO at Sourcegraph and co-creator of AMP, dives into agentic coding tools that enhance developer workflows. He explains AMP's innovative multi-model design, which automates code editing and testing while showcasing strategies for handling model limitations. The conversation touches on the critical importance of terminal UI for a better user experience, along with the concept of 'agent flow' for structured development workflows. Beyang also shares practical insights for skeptics and the sustainable pricing model behind AMP's technology.
AI Snips
Chapters
Transcript
Episode notes
Model Selection Should Be Invisible
- AMP abstracts model selection so users don't pick specific LLMs manually.
- The product picks the best model per subtask to optimize latency and competency.
Always Review Generated Code
- Read and review agent-generated code; don't trust it blind.
- Keep doing human code review to catch subtle bugs and complexity the agent may introduce.
Learn Prompting As A Skill
- Invest time to learn prompting; it's a skill like any other.
- Develop intuition for what agents can and cannot do to get reliable results.