Discovery 001 — PageMotor Is More Than a CMS

Our first architectural question: What problem is PageMotor trying to solve beyond ordinary content management?

Question

Is PageMotor best understood as a traditional CMS with AI features, or as an AI-capability platform that also manages content?

Why This Matters

The answer changes how plugins should be designed. In a conventional CMS, a plugin is usually framed as a website feature. In an AI-first system, a plugin may instead define a new capability, vocabulary, workflow, interface, or action that an authorized agent can use.

Current Hypothesis

PageMotor is more than a traditional CMS. Its architecture appears to organize content, themes, plugins, documentation, AI agents, permissions, and callable actions into a system intended for both human and AI operation.

Confidence: Moderate. The available API structure supports the hypothesis, but the deeper design intent and plugin mechanics still require direct study.

Evidence Available So Far

  • The API exposes a registry of callable actions grouped by controllers, plugins, and themes.
  • Plugins can expose extension actions in addition to core actions.
  • A Valet registry documents methods that plugins or themes can implement to add capabilities.
  • The site reports dedicated AI configuration and AI-agent functions.
  • PageMotor provides guided MCP skills for plugin creation and design conversion.
  • Theme behavior is exposed through structured operations for templates, blocks, instances, design data, CSS, and placement.

Working Mental Model

User request
    ↓
AI agent or API client
    ↓
Available actions and guided skills
    ↓
Controllers, plugins, themes, and Valets
    ↓
Content, design, settings, users, or external behavior

Verified Facts

  • PageMotor manages conventional website content.
  • It also exposes structured, permission-aware actions for programmatic operation.
  • Plugins and themes can participate in the capability system.
  • The platform includes AI-specific configuration and guided MCP skills.

What Is Not Yet Verified

  • Whether the PageMotor designers formally describe the platform as an “AI-capability platform.”
  • Whether every plugin is intended to expose AI-usable behavior.
  • How much reasoning belongs to Skills versus plugin code.
  • Whether Valets are merely registration hooks or a broader capability abstraction.

Design Implications

Until disproven, plugin design should begin with a capability statement rather than a feature list:

After this plugin is installed, what new sentence should an authorized user or AI agent be able to say and have PageMotor understand?

Open Questions

  • What exactly is a PageMotor Plugin?
  • What exactly is a Valet?
  • What distinguishes a Skill from an Action?
  • How are plugin capabilities discovered and permissioned?
  • Which parts of a plugin are visible to an AI agent?

Next Discovery

Discovery 002 — The Controller and Action Architecture.