Rubric Schema-first structured extraction from LLMs. You define a Zod schema. Rubric puts that schema on the request, pulls JSON out of the reply, validates it, and reasks with the error until the value conforms — or retries run out. For documents longer than the context window, splits the text, extracts per chunk, and merges — measured to recover 100% of planted entities at a 0% duplicate rate where a single call over the same document returns nothing (benchmarks). does not patch the SDK. The original client is unchanged.
| Stars | 88 |
| Forks | 6 |
| Language | TypeScript |
| Category | AI Tool |
| License | MIT |
| Quality Score | 64.3741297085231/100 |
| Last Updated | 2026-09-28 |
| Created | 2026-09-08 |
| Platforms | node |
| Est. Tokens | ~20k |
These tools work well together with rubric for enhanced workflows:
Explore other popular ai tool tools:
rubric is Schema-first structured extraction from LLMs — validated JSON from one call or a 150-page document.. It is categorized as a AI Tool with 88 GitHub stars.
rubric is primarily written in TypeScript. It covers topics such as anthropic, data-extraction, document-processing.
You can find installation instructions and usage details in the rubric GitHub repository at github.com/tomnio/rubric. The project has 88 stars and 6 forks, indicating an active community.
rubric is released under the MIT license, making it free to use and modify according to the license terms.