Internal Use

KEYWORD
OWNERSHIP.

Every core query should have one primary landing page. Supporting pages exist to deepen the cluster, not compete for the same click.

TopicKeywordPrimary PageSupporting PagesIntent & Notes
System Prompt Architectureopenclaw 9-layer system prompt/blog/openclaw-agent-system-prompt-architecture-9-layers

Understand the canonical OpenClaw prompt structure.

Keep one main ranking URL and use secondary pages only for supplementary framing.

System Prompt Architectureopenclaw system prompt architecture/topics/system-prompt-architecture

Find the high-level architecture and layer responsibilities.

The hub owns broader cluster intent; the article owns exact architecture breakdown intent.

System Prompt Architectureai agent prompt architecture/blog/engineering-better-ai-agent-prompts-with-software-design-principles

Compare design patterns for complex prompt systems.

Use supporting content for methodology, not duplicate architecture titles.

OpenClaw Nodes & Automationopenclaw node tutorial/blog/openclaw-node-tutorial

Learn the main OpenClaw Nodes workflow and setup path.

Keep the tutorial page focused on setup and control patterns; let the hub own the broader cluster.

OpenClaw Nodes & Automationagent-browser electron automation/blog/agent-browser-electron-app-automation-via-cdp

Find the best page for Electron automation with CDP.

Use the secondary article and skill page for implementation paths, not competing titles.

OpenClaw Nodes & Automationopenclaw browser automation/topics/openclaw-nodes

Evaluate how OpenClaw handles browser and desktop control.

The hub should be the broad comparison entry point.

Claude Code Workflowsclaude code autonomous workflow/blog/claude-code-ai-can-now-work-independently

Understand what autonomous Claude Code execution can do.

Let the article own the announcement/explainer intent and the hub own broader discovery intent.

Claude Code Workflowsclaude code long-running tasks/blog/manus-integration-enables-long-task-capability-in-claude-code

Find the best path to durable or extended task execution.

Keep Manus-specific queries on the Manus integration page.

Claude Code Workflowsclaude code task management/blog/using-linear-as-ai-task-management-hub

See how to operationalize Claude Code work across multiple tasks.

Task-management intent belongs to the Linear page; don’t reuse that framing elsewhere.

Claude Code Workflowsclaude code productivity stack/topics/claude-code

Compare the broader setup around Claude Code.

The hub owns comparison intent across memory, CLI compression, and workflow design.

AI Agent Workflowsagent orchestration demo/blog/codex-monitor-agent-orchestration-demo

See a concrete orchestration example with multiple agents.

The demo page owns concrete orchestration intent.

AI Agent Workflowsai short video workflow/blog/ai-short-video-factory

Find a real automation pipeline for short-form video output.

Use this page for workflow execution intent, not a generic hub title.

AI Agent Workflowsimage generation skills for agents/blog/building-image-generation-skills-for-ai-agents

Learn how to build reusable image-generation capabilities for agents.

Owns build/tutorial intent for visual skill creation.

AI Agent Workflowsmulti-model agent workflow/blog/multi-model-workflow-optimization-with-ccg

Compare how multiple models collaborate in one workflow.

Keep multi-model comparison intent on the CCG page.