Creative Automation / Foundation

Google OKF + MCP : Explained The New "AI Context Stack"

This video explains the 'AI context stack': Google's OKF (Open Knowledge Format, v0.1 from June) gives agents durable curated knowledge as plain markdown bundles in Git, while Anthropic's MCP gives them live access to real tools and data — and shows how the two click together so an agent reads what a metric means from OKF, then runs the real query through MCP.

Cloud Codes9 minTranscript found

Quick learning frame

Read this before watching.

Creative automation uses agents to accelerate production while keeping human taste in story, pacing, selection, and critique.

New playlist item from Cloud Codes; queued for transcript-backed review, topic mapping, and a practical learning artifact.

Skill you build: The ability to diagnose whether an AI failure is a knowledge gap or an access gap, and to fix each with the right layer — OKF for durable meaning, MCP for live reach.

Watch for the shift from claim to mechanism. The learning value is the point where the transcript reveals a repeatable action, tool boundary, context move, review habit, or artifact.

Concept diagram

Where this video fits.

01Brief

02Source

03Generation

04Selection

05Edit

06Taste Review

Deep lesson

Turn this video into working knowledge.

1,478 cleaned transcript words reviewed across 422 timed caption segments.

Thesis

Google OKF + MCP : Explained The New "AI Context Stack" teaches a practical creative automation move: This video explains the 'AI context stack': Google's OKF (Open Knowledge Format, v0.1 from June) gives agents durable curated knowledge as plain markdown bundles in Git, while Anthropic's MCP gives them live access to real tools and data — and shows how the two click together so an agent reads what a metric means from OKF, then runs the real query through MCP.

The goal is not to remember the video. The goal is to extract the operating principle, tie it to timestamped evidence, test how far the claim transfers, and make something reusable.

0:14

Two gaps, two standards

“context is really two different things. What the model knows and what it can reach. In the last year, two open standards quietly showed up to fix exactly that. One is called OKF. The other is called MCP.”

A frontier model dropped into your company hits two different walls: a knowledge gap (it doesn't know your schemas, metric definitions, or join paths, which live scattered across catalogs, wikis, code comments, and engineers' heads) and an access gap (even when it knows what to do, it can't touch your database, Slack, or issue tracker) — OKF is the agent's memory, MCP is its hands. For one question your AI gets wrong at work, write down which half failed: did it not know what the thing means (knowledge), or could it not reach the live system (access)?

4:41

How each layer works

“model. And a server only ever offers three kinds of things. Tools, actions it can take like sending a message or running a query, resources, data it can read like files and database rows and prompts, reusable templates,...”

OKF is 'aggressively boring' — a bundle is just a directory of markdown files, one concept per file with a small YAML block where only 'type' is required and a 'resource' URI pointing at the real live table or API, with ordinary markdown links turning the folder into a knowledge graph — while MCP has three roles (host, client, server) and three primitives (tools, resources, prompts), was adopted by OpenAI, Google, and Microsoft, hit 10,000+ public servers and 97 million monthly downloads, and now lives at the Linux Foundation. Write one OKF-style concept file for a metric you own: type and title in YAML front matter, a plain-English definition, the exact join, and a resource URI to the real table.

6:51

Stack them together

“own, a model invents what your metric means. Give it MCP alone and it runs live queries against a schema it does not understand. You need both. OKF makes the answer correct. MCP makes it current. Together, you...”

OKF is the slow layer (curated knowledge changing by pull request in Git) and MCP the fast layer (live access changing by the second); the handoff is one motion — the agent opens the OKF concept for 'weekly active users', grabs the definition and resource URI, then MCP runs the real query — and since MCP can serve resources, you can expose the whole OKF bundle through an MCP server, while the economics collapse 10x10=100 custom connectors into 10+10. Trace one real question ('what were weekly active users last week?') through the stack on paper: which file the agent reads, what URI it extracts, and which MCP tool call executes the query.

01

Brief

Start with this video's job: This video explains the 'AI context stack': Google's OKF (Open Knowledge Format, v0.1 from June) gives agents durable curated knowledge as plain markdown bundles in Git, while Anthropic's MCP gives them live access to real tools and data — and shows how the two click together so an agent reads what a metric means from OKF, then runs the real query through MCP. Treat "Brief" as the outcome you are trying to make visible, not a topic label. Anchor it to 0:14, where the video says: “context is really two different things. What the model knows and what it can reach. In the last year, two open standards quietly showed up to fix exactly that. One is called OKF. The other is called MCP.”

02

Source

Use "Source" to locate the part of the creative automation workflow the video is demonstrating. Ask what changes in your real setup if this claim is true. Anchor it to 4:41, where the video says: “model. And a server only ever offers three kinds of things. Tools, actions it can take like sending a message or running a query, resources, data it can read like files and database rows and prompts, reusable templates,...”

03

Generation

Turn "Generation" into the reusable artifact for this lesson: A creative workflow board with critique criteria and review checkpoints. This is where watching becomes something you can inspect and reuse.

04

Selection

Use "Selection" as the application surface. Decide whether the idea touches a browser flow, a local file, a model choice, a source document, a UI, or a review step.

05

Edit

Use "Edit" to prove the lesson. The evidence should connect back to the video title, transcript anchors, and a concrete output, not a generic best-practice claim.

06

Taste Review

Use "Taste Review" to carry the idea forward: save the prompt, checklist, diagram, or operating rule that would make the next agent run better.

Example

Source-backed work packet

Convert the video into a scoped task that includes the transcript claim, target workflow, acceptance criteria, and proof. The output should be a creative workflow board with critique criteria and review checkpoints..

Example

Claim vs. demo brief

Separate what the speaker claims, what the demo actually proves, and what still needs outside verification before you adopt the workflow.

Example

Teach-back module

Transform the lesson into a definition, a mechanism diagram, one misconception, one practice exercise, and a check-for-understanding question.

Do not learn it wrong

Treating the title as the lesson without checking what the transcript actually says.
Letting the prompt drift into generic advice that could apply to any video in the playlist.
Copying the tool setup without identifying the operating principle that transfers to your own stack.
Skipping the artifact, which means the learning never becomes operational or inspectable.

Transcript-derived moments

Use timestamps to study the actual video.

Problem frame

“context is really two different things. What the model knows and what it can reach. In the last year, two open standards quietly showed up to fix exactly that. One is called OKF. The other is called MCP.”

Working mechanism

“model. And a server only ever offers three kinds of things. Tools, actions it can take like sending a message or running a query, resources, data it can read like files and database rows and prompts, reusable templates,...”

Transfer moment

“own, a model invents what your metric means. Give it MCP alone and it runs live queries against a schema it does not understand. You need both. OKF makes the answer correct. MCP makes it current. Together, you...”

Quality check

Do not count this as learned until these are true.

01

State the transcript-backed claim in your own words: This video explains the 'AI context stack': Google's OKF (Open Knowledge Format, v0.1 from June) gives agents durable curated knowledge as plain markdown bundles in Git, while Anthropic's MCP gives them live access to real tools and data — and shows how the two click together so an agent reads what a metric means from OKF, then runs the real query through MCP.

02

Explain the practical stakes without hype: New playlist item from Cloud Codes; queued for transcript-backed review, topic mapping, and a practical learning artifact.

03

Map the idea onto the Brief -> Source -> Generation -> Selection -> Edit -> Taste Review sequence and name the weakest link.

04

Produce the artifact and include the evidence that proves it: A creative workflow board with critique criteria and review checkpoints.

Put it into practice

Give this grounded prompt to Codex or Claude after watching.

You are helping me turn one specific YouTube video into real, durable learning.

Source video:
- Title: Google OKF + MCP : Explained The New "AI Context Stack"
- URL: https://www.youtube.com/watch?v=bwQ70NMd57k
- Topic: Creative Automation
- My current learning frame: Pick one metric your team argues about, write its OKF concept file with the definition, join path, and resource URI, then connect an MCP server to the warehouse and verify the agent answers correctly by reading the file first and querying live second.
- Why this matters: New playlist item from Cloud Codes; queued for transcript-backed review, topic mapping, and a practical learning artifact.

Transcript anchors from this exact video:
- 0:14 / Evidence 1: "context is really two different things. What the model knows and what it can reach. In the last year, two open standards quietly showed up to fix exactly that. One is called OKF. The other is called MCP."
- 1:53 / Evidence 2: "an access gap. Even when it does understand, it cannot actually touch anything. Two different problems and for years we solved them badly. Take the knowledge gap first. The truth about your systems is scattered everywhere. A metadata..."
- 4:41 / Evidence 3: "model. And a server only ever offers three kinds of things. Tools, actions it can take like sending a message or running a query, resources, data it can read like files and database rows and prompts, reusable templates,..."
- 6:51 / Evidence 4: "own, a model invents what your metric means. Give it MCP alone and it runs live queries against a schema it does not understand. You need both. OKF makes the answer correct. MCP makes it current. Together, you..."

Your task:
1. Use the transcript anchors above as the primary source packet. If you add outside context, label it clearly as outside context and keep it secondary.
2. Create a source-check table with columns: timestamp, claim, what the demo proves, confidence, and what still needs verification.
3. Extract the actual teachable claims from the video. Do not invent claims that are not supported by the title, lesson frame, or transcript anchors.
4. Build a reusable learning artifact: A creative workflow board with critique criteria and review checkpoints.
5. Include:
- a plain-English definition of the core idea
- a diagram or structured model using this sequence: Brief -> Source -> Generation -> Selection -> Edit -> Taste Review
- 3 concrete examples that apply the video idea to real agentic work
- 2 failure modes the video helps prevent
- a checklist I can use the next time I run Codex or Claude
- one practical exercise with a clear done signal
6. Add a "learning transfer" section: what changes in my workflow tomorrow if I actually learned this?
7. Add a "source check" section that cites which transcript anchor supports each major takeaway.

Quality bar:
- Make this specific to "Google OKF + MCP : Explained The New "AI Context Stack"", not a generic Creative Automation essay.
- Prefer operational examples, failure modes, and reusable artifacts over broad definitions.
- Call out uncertainty instead of smoothing over weak evidence.
- If evidence is weak, say what transcript segment or timestamp needs review instead of guessing.
- Finish with a concise artifact I could paste into my learning app.

Misconceptions

What to stop believing.

Creative AI removes the need for taste.

It increases the need for taste because output volume explodes.

The best prompt is enough.

References, critique, iteration, and post-production matter just as much.

Practice studio

Learning only counts when you make something.

01

Transcript evidence map

Separate what the video actually says from what you already believe about the topic.

3 source-backed takeaways with timestamps, confidence, and a transfer note.

02

One useful artifact

Apply the video to a real workflow and produce a creative workflow board with critique criteria and review checkpoints..

A reusable artifact with a done signal and one verification step.

03

Teach-back card

Explain the lesson to someone who has not watched the video yet.

A 90-second explanation, one diagram, one example, and one misconception to avoid.

Recall check

Answer first, then reveal — without rewatching.

What are the two halves of the context problem, and which standard solves each?

In an OKF concept file, which field is required and what does the 'resource' field do?

Why do you need both OKF and MCP rather than either alone?

Source shelf

Use the video as a doorway, then verify with primary sources.

ReadingComfyUIwww.comfy.org/ReadingAffinityaffinity.serif.com/