CRD

Learning / guidance

Learn Capability Documentation

Understand why CRD exists, where the Capability MLE idea comes from, and how capabilities connect product intent to software, agents, and operations.

Start with Why CRD? →Read the normative specification →

These pages explain and interpret the framework. They do not add requirements to a CRD. The CRD Specification remains the normative source.

Start here

Why capability-centered documentation?

Why CRD?

See why a stable capability layer matters as implementations become fluid across software and agents.

Read →

Watch the overview

Start with the 9-minute Capability Requirements video, then continue through the guided media sequence.

Watch →

Foundations

Find the right semantic boundary.

Minimum Logical Element

Learn the scope discipline: minimum size while retaining contextual logical meaning.

Explore →

From features to capabilities

Compare CRDs with features, PRDs, Storybook, APIs, MCP, skills, and tools.

Compare →

Apply it

Turn meaning into usable documentation.

Transform existing material

Guide an agent through identify, combine, decompose, classify, trace, and document.

Apply →

Preserve shared source context

Keep product-wide constraints and intent visible when they belong to no single capability.

Use →

Use the CRD Author skill

Give an installable agent the Define and Extract workflows, including inventory and source-context handling.

Open on GitHub →

Study examples

See customer-facing and internal Capability MLEs with contracts and realizations.

Explore →

Read, watch, listen

Deepen the context.

Why AI agents break documentation

Listen to the 35-minute exploration of the documentation problem agents expose.

Listen →

The Semantic Blueprint

Read the 15-page conceptual resource, then watch the architecture-focused follow-up.

Read →