Chrome DevTools MCP shipped with one tool: debug_webpage. Agents failed silently because they couldn't compose behaviors. The team decomposed it into 25 focused tools and assumed the problem was solved. It wasn't — now agents had 25 tools and no reliable way to pick the right one. Michael Hablich's talk is an honest account of building the same thing wrong three times and what the fixes actually looked like.

The concrete lessons: semantic summaries instead of raw 50,000 line JSON trace files, error messages rewritten so agents can self heal without a human in the loop ("Cannot navigate back, no previous page in history" instead of "Unable to navigate back in currently selected page"), a metric called tokens per successful outcome to measure interface fuel efficiency, and a deliberate decision to keep the autoconnect friction rather than remove it once they thought through prompt injection and the lethal trifecta.

Speaker info:
- https://x.com/MHablich
- https://www.linkedin.com/in/michael-hablich/

AI Engineer

Michael Hablich, Chrome DevTools PM at Google, shares four engineering lessons from building Chrome DevTools for AI agents via MCP. Key insights include: treating agents as a distinct user class with different cognitive bottlenecks than humans; measuring interface efficiency using 'tokens per successful outcome'; returning semantic summaries instead of raw data to avoid blowing context windows; decomposing monolithic tools into granular ones with clear descriptions; implementing error recovery playbooks and proactive detours for agent self-healing; and maintaining deliberate trust boundaries (friction by design) to prevent prompt injection and unauthorized access across three tiers of agent deployment environments.

Building Agent Interfaces: Lessons from Chrome DevTools (MCP) for Agents — Michael Hablich, Google