Panel: Agentic SE & Open Source in the FMware Era – Toward Trustworthy, Context-Aware FMware
summary
Panel argues coding agents (Google jewels, OpenAI codeex) scale code generation faster than human review and demands new IDEs, observability and cost tradeoffs; teams build claw.md mentorship files to steer agents, run 10-agent ensemble runs to generate diverse PRs, and track asset provenance (agent, model, params) because traditional VCS lacks context and inference metadata; warns of prompt injection, runaway agent loops burning API budgets, and the need for domain-specific benchmarks and freshness checks before shipping.
AI-written summary · about
chapters (12)
- 0:00Panel introduction
- 0:21Coding agents challenges
- 3:16LLM integration tooling
- 7:37Advice for junior devs
- 12:39100x developer skills
- 16:15Prompt engineering mastery
- 19:26LLM selection guide
- 24:30Creativity vs automation
- 30:53Secure AI deployment
- 37:33Mentoring AI agents
- 42:06Developer roles future
- 57:29Closing & Q&A
related talks
46mPanel Discussion: The Agentic Software Factory That Enterprises Actually Need
50mPanel: Beyond The Vibe: Code Quality First
50mPanel Software Engineering in the age of AI
38mCoding Agents & Language Evolution: Navigating Uncharted Waters
25mBeyond the IDE: Toward Multi-Agent Orchestration
28mXtremeJS Xperts Panel: The AI Revolution in Coding – Tools, Trends, and Tomorrow
32mProgramming language evolution in the AI era
51mFuture of Software Development