| name | output-scoring-en |
| description | Scores the quality of an AI legal output before it goes out, in two layers: objective (mechanically checkable - do the citations exist, are the provisions and dockets real) and subjective (LLM-as-judge on a 1-5 rubric: legal correctness, completeness, clarity, jurisdiction fit, grounding/anti-hallucination). Returns a scorecard and a decision: send / revise / full verification. A layer above the deliverable - it does not write, it scores. Jurisdiction-neutral. Pairs with reviewer-en, adversarial-legal-review-en and legal-syllogism-en. Use when: "score this output", "is this ready to send", "quality of this legal answer", "scorecard for the deliverable", "rubric check" - before sending an opinion, memo, or brief.
|
| license | Apache-2.0 |
| allowed-tools | ["Read"] |
| data-residency | local |
| requires-human-approval | false |
| pii-egress | none |
| attribution | {"source":"FudanDISC/DISC-LawLLM (Fudan DISC Lab)","url":"https://github.com/FudanDISC/DISC-LawLLM","license":"Apache-2.0","relationship":"clean-room","note":"Two-layer method (objective matching plus a subjective LLM-as-judge rubric on a 1-5 scale) based on DISC-Law-Eval. Exam data and judge model dropped; rubric, dimensions and thresholds written from scratch. English counterpart of ocena-outputu-pl.\n"} |
| metadata | {"canonical_source":"https://github.com/matematicsolutions/awesome-matematic-skills-en/blob/main/verification-foundation/skills/output-scoring-en/SKILL.md - the maintained version. Catalogue copies are snapshots and may be out of date; check the canonical file before relying on any legal reference in this skill.\n","author":"Wiesław Mazur / MateMatic","version":"1.0.0","companion_skills":"reviewer-en, adversarial-legal-review-en, legal-syllogism-en"} |
Output scoring - a scorecard before you send
Philosophy
Before you send it, score it - and keep what is checkable apart from what must be judged. An AI
output can read smoothly and still be wrong. This skill scores it in two layers: objective (facts you
can verify mechanically) and subjective (quality you have to judge against an explicit rubric).
Mixing the two is where false confidence comes from - "sounds right" is not "the citations check out".
It scores; it does not write or send. The decision is the lawyer's; the scorecard is the basis.
Layer 1 - objective (mechanically checkable)
Delegate, do not guess:
- Citations and provisions - do they exist in the source verbatim; were they all caught.
- Dockets / identifiers - real and correctly written.
- Completeness - did any RED finding drop out of the summary.
The objective result is binary per item: matches / does not. Even one mismatched citation means the
output is not ready, whatever the subjective score.
Layer 2 - subjective (1-5 rubric, LLM-as-judge)
Score each dimension 1-5 with a one-sentence reason:
| Dimension | 1 | 5 |
|---|
| Legal correctness | thesis wrong or unsupported | thesis sound and grounded in the rule |
| Completeness | omits material issues | covers everything material to the question |
| Clarity | muddled, inconsistent | precise, readable for the reader |
| Jurisdiction fit | confuses systems/rules | correct for the governing law |
| Grounding (anti-hallucination) | claims with no basis | every claim has a basis |
Decision
- Send - Layer 1 clean and subjective average >= 4, no dimension < 3.
- Revise - Layer 1 clean but a dimension scores 2-3 (name it).
- Full verification - Layer 1 has a mismatch or a dimension = 1 -> route to grounding / adversarial review.
Output format
LAYER 1 (objective): citations OK/X | provisions OK/X | dockets OK/X | completeness OK/X
LAYER 2 (1-5 rubric):
Legal correctness: 4 - <reason>
Completeness: 3 - <...>
Clarity: 5 - <...>
Jurisdiction fit: 4 - <...>
Grounding: 4 - <...>
Average: 4.0
DECISION: Revise (Completeness 3 - limitation period not addressed)