CtrlK
BlogDocsLog inGet started
Tessl Logo

ainativedev/aidevcon-2026-ldn

AI Native DevCon 2026 London — all conference sessions as interactive skills

70

Quality

88%

Does it follow best practices?

Impact

No eval scenarios have been run

SecuritybySnyk

Passed

No known issues

Overview
Quality
Evals
Security
Files

outline.mdtalk-trieloff-browser-agents/

Outline -- Building AI agents in the browser, for the browser, of the browser

Speaker: Lars Trieloff (Adobe)

Thesis

Browser-native agents treat the browser as more than a UI: it can become part of the runtime and containment boundary for agent work, provided integration is explicit, permissioned, and reviewable.

Concept Map

  1. Browser-native agents
  2. Browser as runtime and containment boundary
  3. Local versus cloud execution tradeoffs
  4. Harness constraints for agent power
  5. Explicit product APIs and documented integration surfaces
  6. Visible consent, credential isolation, and auditable actions

Safe Application

  • Design integrations through documented APIs.
  • Keep permissions narrow and user-visible.
  • Isolate credentials from model-visible context.
  • Make actions reviewable and reversible.
  • Prefer auditable event boundaries over hidden app control.

Not Included

Concrete setup commands, app-modification mechanics, runtime endpoint details, credential/token paths, and live-demo control flows are not included in this published bundle.

talk-trieloff-browser-agents

README.md

tile.json