Output Scoring

Score the quality of an AI-produced legal output before it goes out, in two layers kept deliberately apart: an objective layer of mechanically checkable facts (do the cited authorities exist, are the provisions and docket numbers real, did any red-flagged finding drop out of the summary) and a subjective layer judged against an explicit 1-5 rubric (legal correctness, completeness, clarity, jurisdiction fit, grounding). Returns a scorecard and one of three decisions: send, revise, or route to full verification. A layer above the deliverable - it scores, it does not write or fix. Jurisdiction-neutral. Use before sending an opinion, memo, or brief, or whenever the question is whether an AI-assisted draft is ready to go out.

open-legal-products Updated 97 repo stars

File contents

open-legal-products/mike-workflows commit 379f6e6d1c

Frequently asked questions

npx get-skill add open-legal-products/output-scoring