Starter packets are practical, source-backed materials that help people and supervised agents work faster: project instructions, agent workflows, policy references, evaluation rubrics, integration guides, public-source digests, and structured research references.

Rubrics, checklists, scoring guides, and review templates for evaluating outputs, datasets, prompts, workflows, or product surfaces.

Starter AI/workflow packet categories, collection pages, Studio helper copy, and agent-readable listing filters are available for reviewed listings. Publication remains review-gated; empty collections may appear until supply is approved.

Good evaluation rubric packets state what is being evaluated, what evidence matters, how scores or verdicts are assigned, and which claims are judgment-based. Graunt uses evaluated quality language for these materials unless a claim is backed by objective proof.