Product Structure

The antora-tracer package ships two independent capabilities that share one syntax. This page draws the boundary between them so readers and coding agents can tell core from add-on.

Core: the traceability extension

The core is the Antora extension and its CLI. It parses [item] blocks, builds an in-memory graph, and generates matrices, visualizations, and exports.

The core lives in src/ and is documented in the Reference section and on the Architecture page. It works entirely without the agent skills.

Add-on: the agent skills

The add-on is a set of Agent Skills that teach an AI assistant how to write and review traceable content. The package ships three:

  • requirements-writing — create and review functional requirements in EARS style

  • use-case-engineering — create and review use case descriptions

  • review-architecture-diagrams — review architecture diagrams

The skills live in skills/ and are documented in the install and evaluate how-to guides. They work independently of the extension — a skill guides an agent to produce correct [item] blocks, but the extension is not required to run the skill.

What they share

Both capabilities use the same [item] block syntax (the item macro) and the same traceability.yml configuration model. A skill produces content the extension consumes, but neither depends on the other.

Naming note

The requirements module in the example site is not the requirements-writing skill. The module holds the project’s own requirements specification — REQ items produced as self-traceability data. The skill is a product feature that teaches an agent to write such requirements. The two share a name but are unrelated.

Repository map

src/       core extension and CLI
skills/    add-on agent skills
evals/     deterministic and LLM-based skill testing
examples/tracer/modules/requirements/   the project's own requirements spec (data, not product docs)