Managing fleets of coding agents with OpenHands
summary
Demonstrates OpenHands (MIT-licensed) running as a CLI or web app to launch cloud or local sandboxes, use a Trivy scanner to detect vulnerabilities, then spawn one agent per CVE; the live demo finds three vulnerabilities and opens three PRs. Claims orchestration gets you to roughly 90% automation and 10% human review, recommends about 3 to 5 concurrent agents before complexity spikes, and notes agents run anywhere (Docker/Kubernetes), handle merge conflicts, and work with small models from ~3B up to 32B.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:12Evolution of AI development
- 2:43Agent orchestration overview
- 4:35Practical use cases
- 6:15Workflows and methodology
- 7:25Implementation strategies
- 9:48CVE remediation demo
- 15:35Q&A and conclusion
related talks
46mVibecoding your Own Multi-Agent Workstation
17mSoftware Development Agents: What Works and What Doesn't
24mAI Agent Swarms Are Amazing
01h 16mAutomating Large Scale Refactors with Parallel Agents
18mCollaborative AI Engineering: One Dev, Two Dozen Agents, Zero Alignment
15mDevelop at Idea Velocity
16mWhat Comes After AI Coding Assistants
40mHow To Tame Your Agents?