---
title: "Slides: Why MCP and ChatGPT Apps Use Double Iframes — Frédéric Barthelet, Alpic"
category: "slides"
video_id: "c-2eEv2ou7Y"
sourceLabels: ["Public YouTube video frames", "Public YouTube metadata"]
---

# Slides: Why MCP and ChatGPT Apps Use Double Iframes — Frédéric Barthelet, Alpic

## Source Video
[Why MCP and ChatGPT Apps Use Double Iframes — Frédéric Barthelet, Alpic](https://www.youtube.com/watch?v=c-2eEv2ou7Y)

## Relationship To World's Fair 2026
These slides are extracted from a public AI Engineer YouTube video connected to World's Fair 2026. Speaker-matched clips are supporting context unless later confirmed as exact session recordings; official livestream recordings are day-level/event-level source material.

## Related Scheduled Sessions
- No individual scheduled session mapping has been assigned yet; treat this as an event livestream deck.

## Extracted Slides
![[assets/slides/c-2eEv2ou7Y/slide-001.jpg]]

OCR text:

> _ a |
> a in - om ie ce | Why MCP & ChatGF
> | ammm +--—. ell Double Iframes
> ‘| | em Pi TES i 4 g ‘ z
> ae

![[assets/slides/c-2eEv2ou7Y/slide-002.jpg]]

OCR text:

> N
> a
> Mae alg ?
> Pee G Pe are
> * a a anne
> Sg a Tere lalM eel CMO mOUr ICCA mile Rael eels ae
> * red * inside Al chats FO
> Se
> Liad and Ido experimenting with MCP-UI a
> Initially released on ChatGPT as OpenAl's Apps SDK & aa
> Standardized as MCP first official extension or ae :
> nos Ea
> ry coed ar
> : |
> Engineering the future of Al

![[assets/slides/c-2eEv2ou7Y/slide-003.jpg]]

OCR text:

> N
> . arenas
> Sein Cents
> i CUs Sana cd
> a
> bf * eee ras Oe
> bd bf <
> * * ip
> oe “a iz beter ti
> ry FY Views render in response to tool calls 7 *
> * Roe at ar)
> , c
> HTML documents live in resources Sa ee eee err
> Py
> Views can be cached ahead of time sr te
> rene Ae oe rer ee
> «
> a Or ee ae
> | teil ee aooat
> a)
> . AIEngi
> SUL se) a

![[assets/slides/c-2eEv2ou7Y/slide-004.jpg]]

OCR text:

> NJ
> NY
> Before AI Apps —
> lus tent Sevurtty Solis tte rated Em aY
> defaglt-sre
> i ee eefauit-
> * bf Peer ereas [ none’ J ‘sett’ Frets: |
> * * re wc
> aa es oa
> bak ae
> bd
> RSC “sett T cunsate-intine: J ‘nonce ]
> UM 31! nttos: Puss:
> SMM ‘sett’ J data: | plod:
> fone ED I OS
> a |
> Engineering the future of Al

![[assets/slides/c-2eEv2ou7Y/slide-005.jpg]]

OCR text:

> N ;
> Rendering external HTML?
> Pe
> 5d od
> ‘ ,
> * Ld
> ae ur HTML etement represents a nested browsing context,
> embedding another HTML page into the current one.
> . | PNP sare ass |
> Svea

![[assets/slides/c-2eEv2ou7Y/slide-006.jpg]]

OCR text:

> SS ;
> What if — rs
> inherits parent CSP without script nonce,
> Engineering the future of Al

![[assets/slides/c-2eEv2ou7Y/slide-007.jpg]]

OCR text:

> . oo 4
> oy
> roar ee ee ome Al Engineer
> oo me i ell
> CL eee) Ce
> ‘| Wa inherits parent CSP without script n
> J
> an

![[assets/slides/c-2eEv2ou7Y/slide-008.jpg]]

OCR text:

> pp-
> Thesolution-doubleiframe
> specific subdomain with Cs
> chatgpt.com Content-Security-Policy
> script-srcnonce--https://*,chatgpt.com
> AIE
> frae-aroself·.stripe.com*-youtube.com
> cbody>
> srca*https://abo123.web-sandbox.oalusercontent.com>
> <meta http-equiv='Content-Security-Policy'content=connect-src'sy-servic
> <script>initializeApp(）</scriptx
> ATp/>ddyJno<ATp
> /body
> AlEngineer
> EUROPE

![[assets/slides/c-2eEv2ou7Y/slide-009.jpg]]

OCR text:

> NJ
> 7 f t
> Pel
> bd bd
> ace
> * *. Bu ike) fs Tents) ed
> aed
> L connect Domains | Raa
> [scr ipt—sre [ wedia-sec | Pa oe ms
> een eT on reer)
> een
> 4 os
> eas
> a
> | Engi ing the fut eA

![[assets/slides/c-2eEv2ou7Y/slide-010.jpg]]

OCR text:

> Build MCP Apps with
> ae i
> = Ore @ SKYBRIDGE
> * * bal wy 7 mu Open-source framework for building MCP Apps
> 7 Parad * hi Le rae Tt
> eee
> Or ar a
> & Star the repo mur ela aeRO man EL
> Al Engineer
> aU cela

![[assets/slides/c-2eEv2ou7Y/slide-011.jpg]]

OCR text:

> Staybridge wg wae GEE Lena eae ames Serdeioes *
> tans magic 3.0011 cc meocore, (EEE
> magic ¥ batt . WES. %
> x * &
> bd *
> ace
> bd *
> aan
> ia
> a |
> | Al Engineer |
> au) cela


## Slide-Derived Subjects To Review
Subject extraction uses video title, related session titles/descriptions, transcript context, and OCR text when available. OCR is best-effort and should be reviewed against the embedded slide images.