Agentic AI Atlasby a5c.ai
OverviewWikiGraphFor AgentsEdgesSearchWorkspace
/
GitHubDocsDiscord
i.3Wiki
Agentic AI Atlas · Placement Checklist
agent-generate/universal-agentic-stack/03-placement-checklista5c.ai
Search the atlas/
Wiki · linked records

Article and nearby pages

I.Current articlepp. 1 - 1
Universal Agentic Stack OrientationBuilder and Product FitUniversal Agentic Stack Landscape DiagramWorked ExamplesCommon ConfusionsReading Paths
II.Documented nodesrefs · 8
Model · LayerProvider · LayerTransport · LayerAgent-Core · LayerAgent-Runtime · LayerAgent-Platform · LayerWorkspace · LayerExecution · Layer
I.
Wiki article

agent-generate/universal-agentic-stack/03-placement-checklist

Reading · 2 min

Placement Checklist reference

Use this checklist when someone says "this tool is an agent platform" or "this framework does everything." It forces the classification into concrete responsibilities instead of labels.

Page nodewiki/agent-generate/universal-agentic-stack/03-placement-checklist.mdNearby pages · 11Documents · 11

Continue reading

Nearby pages in the same section.

Universal Agentic Stack OrientationBuilder and Product FitUniversal Agentic Stack Landscape DiagramWorked ExamplesCommon ConfusionsReading PathsComparison MatrixReview TemplateFAQUniversal Agentic Stack LayersUniversal Agentic Stack Source Map

Documented graph nodes

Records linked directly from this page’s Page node.

Model · LayerProvider · LayerTransport · LayerAgent-Core · LayerAgent-Runtime · LayerAgent-Platform · LayerWorkspace · LayerExecution · Layer

Placement Checklist

Use this checklist when someone says "this tool is an agent platform" or "this framework does everything." It forces the classification into concrete responsibilities instead of labels.

Step 1: Pin down the model side

  • What model family or version is actually being used?
  • Who serves it?
  • What transport or protocol carries requests to it?

This covers Layers 1, 2, and 3.

Step 2: Pin down the agent side

  • Where is the actual decision loop or graph logic?
  • Where do tools, approvals, interrupts, and session state live?
  • Where do installed skills, plugins, channels, and extension contracts live?

This covers Layers 4, 5, and 6.

Step 3: Pin down the operating boundary

  • What workspace does the agent read and write?
  • Where do shell commands, browser runs, or tool actions execute?
  • What policy restricts files, network, secrets, or binaries?

This covers Layers 7, 8, and 9.

Step 4: Pin down the surface

  • What actions are exposed to users or systems?
  • How are results rendered, streamed, or integrated?

This covers Layers 10 and 11.

Step 5: Record what is delegated

Not every product owns every layer. Explicitly note when a layer is delegated to:

  • an IDE
  • a host application
  • a CI runner
  • a cloud control plane
  • a user-managed shell or container

Delegated layers still exist. They just are not owned by the product you are classifying.

Step 6: Write the placement in one sentence

Force yourself to write a short sentence with three parts:

1. main owned layers 2. main delegated layers 3. main exposed surface

If you cannot do that clearly, your classification is still too fuzzy.

Minimal worksheet

Layer bandOwnedExposedDelegatedNotes
1-3 Model supply
4-6 Agent system
7-9 Operating boundary
10-11 Surface

Fill this before arguing about labels.

Output format

When you finish, you should be able to say something like:

This product mainly owns Layers 4-6, delegates 7-9 to the host environment, and exposes Layers 10-11 through a web UI and API.

Follow-up

  • Use `layers.md` to verify disputed boundaries.
  • Use `04-worked-examples.md` if you want to compare your result to common product shapes.

Trail

Wiki
Generated References
Universal Agentic Stack

Placement Checklist

Continue reading

Universal Agentic Stack Orientation
Builder and Product Fit
Universal Agentic Stack Landscape Diagram
Worked Examples
Common Confusions
Reading Paths
Comparison Matrix
Review Template

Page record

Open node ledger

wiki/agent-generate/universal-agentic-stack/03-placement-checklist.md

Documents

Model · Layer
Provider · Layer
Transport · Layer
Agent-Core · Layer
Agent-Runtime · Layer
Agent-Platform · Layer
Workspace · Layer
Execution · Layer