FinePrint. My reviews

What does your code prove?

Review a public GitHub repository against the event’s judging rubric. Find the evidence, the gaps, and your next useful fix.

Reads up to 10 files from the current default-branch commit. Selected public code is sent to Modal for analysis; it is not uploaded to the public Sanity dataset. No code is executed. Uses the shared AI allowance.

What this review looks for

Official rubric · captured 2026-09-23. Evidence prompts below are FinePrint’s guidance.

  1. Meaningful use of Sanity Context and structured content

    Look for content schemas, references and retrieval that change application behavior. A dependency or README claim alone does not demonstrate meaningful use.

  2. Technical implementation and code quality

    Inspect implementation, validation, error handling and meaningful tests. Do not claim tests passed or deployed behavior from source alone.

  3. Use of Knowledge Bases

    Look for actual Knowledge Base retrieval and use of the retrieved entries, with source attribution. Configuration alone is insufficient.

  4. Usability

    Inspect task flow, accessibility, loading and failure states. Runtime usability and visual quality still need a live demonstration.