Agentic Security: Permissions, Provenance, and the Agent Supply Chain
Official Schedule Context
- Date/time: 2026-06-29 · 2:25pm-2:45pm
- Track/room: Security · Track 5
- Speaker(s): Steve Yegge
- Session type/status: sponsor · confirmed
Official Description
As AI agents move from demos into production engineering workflows, the security boundary shifts
from code alone to the permissions, tools, prompts, dependencies, credentials, and orchestration
layers that agents can touch. This talk frames agentic security broadly: least-privilege agent
permissions, sandboxing and capability design, provenance for agent-generated changes, risks in
agent/tool/package supply chains, and practical patterns for keeping autonomous coding and
operational agents auditable and containable.
Related YouTube Video
2026: The Year The IDE Died — Steve Yegge & Gene Kim, Authors, Vibe Coding (speaker-match related prior/adjacent AI Engineer video; captions: English auto-captions).
Transcript Status
Related video transcript availability: English auto-captions. Treat this as supporting context, not a recording of this exact scheduled session unless later confirmed. Not fetched yet.
People
Notes
- Pending transcript synthesis when an official recording or confirmed matching video is available.
Supporting Slides
- youtube 7Dtu2bilcFs slides — extracted from the related public AI Engineer video.
Slide Evidence
- Slide-only cropped deck: youtube 7Dtu2bilcFs dense slides (13 viable slide images).
- Related slide/OCR pages:
- youtube 7Dtu2bilcFs dense slides
- youtube 7Dtu2bilcFs reconstructed slides
- youtube 7Dtu2bilcFs slides
- Slide-derived terms:
software,coding,google,deepmind,engineering,year,died,steve,yegge,gene,engineer,author,researcher,code,models,future,will,watch