인터페이스는 현지화되어 있으며, Skill 소개와 설명 문서는 선택한 언어로 AI가 자동 번역합니다. 코드·명령어·경로·전문 용어는 원문을 유지하며 원작자 문서가 우선합니다.
Skills 둘러보기
디자인 및 창작 · SKILL.md

cross-exam-summary

cross-exam-summary

Generates a structured cross-examination summary from depositions, exhibits, and discovery organized by witness and theme with pinpoint citations. Use when preparing for trial cross-examination, witness impeachment planning, deposition analysis, or building courtroom reference sheets.

Independent GitHub snapshot · Automated classification · Original documentation governs

Skill 문서

GitHub

name: cross-exam-summary language: en description: Generates a structured cross-examination summary from depositions, exhibits, and discovery organized by witness and theme with pinpoint citations. Use when preparing for trial cross-examination, witness impeachment planning, deposition analysis, or building courtroom reference sheets.

Cross-Examination Key Points Summary

Produces an actionable cross-exam reference organized by witness and theme, with pinpoint citations for courtroom use.

Prerequisites

  • Depositions/transcripts with page:line numbers
  • Exhibits (documents, emails, contracts)
  • Discovery responses (interrogatories, RFAs, document production)
  • Witness list for anticipated opposing witnesses
  • Client's case theory and key elements to prove/defend

Quick Start

  1. Identify all witnesses and group testimony by theme (not chronology)
  2. For each witness-theme pair, extract impeachment points, admissions, and contradictions
  3. Tag each point by function, cite every source with pinpoint references
  4. Build the quick-reference table from highest-impact points
  5. Cross-reference exhibits to witnesses and contradictions

Output Structure

Quick-Reference Table (Top of Document)

WitnessTop Impeachment PointSourceTheme
{Name}{One-line summary}{Doc, p.XX:LL-LL}{Theme}

List 2-3 highest-impact points per witness.

Per-Witness Sections

Organize each witness by theme:

## [Witness Name] — [Role]

### Theme: [e.g., "Knowledge of Defect"]

**Purpose:** [IMPEACH | ADMIT | FOUNDATION | CREDIBILITY | ELEMENT | EXPERT]

- **Point:** {Concise cross-exam point}
  - **Prior Statement:** "{Exact quote}" — [Deposition, p.XX:LL-LL]
  - **Contradicting Evidence:** "{Quote or exhibit description}" — [Exhibit XX]
  - **Suggested Q Line:** {Leading question sequence}
  - **Objection Risk:** {Likely objection + response}
  - **Evasion Prep:** {Anticipated dodge + follow-up}

Point Tags

TagPurpose
IMPEACHPrior inconsistent statement (FRE 613)
ADMITElicit favorable admission
FOUNDATIONEstablish/undermine foundation for testimony
CREDIBILITYBias, interest, motive, perception, memory
ELEMENTDirectly proves/negates a claim element
EXPERTChallenge methodology, assumptions, qualifications

Expert Witnesses

For experts, additionally address:

  • Methodology challenges (Daubert/Frye factors)
  • Unsupported or contradicted factual assumptions
  • Qualification gaps relevant to opinions offered
  • Prior inconsistent opinions in other cases

Exhibit Cross-Reference Table

Exhibit #DescriptionWitnessesKey Contradiction
{Ex. #}{Brief desc}{Names + cite}{What it undermines}

Rules

  • Citation format: [Document Name, p.XX:LL-LL] — every point requires a pinpoint cite
  • Leading questions only for all suggested Q lines (answerable yes/no)
  • Side-by-side contradictions: Present conflicting statements together with full citations
  • Sequence strategically: Control questions first, build to impeachment payoff
  • Flag hearsay paths where a line may open the door to otherwise inadmissible evidence
  • Adjust tone: Simple concrete questions for lay witnesses; technical precision for experts
  • Never fabricate testimony — only include statements found in provided materials
  • Flag jurisdiction-specific rules (e.g., Queen Caroline's Rule states requiring foundation before extrinsic evidence)

Key changes from original:

  • Removed tags — not part of the Agent Skills spec (only name and description in frontmatter)
  • Tightened description — shorter, still includes clear trigger guidance
  • Added Quick Start — concise 5-step workflow for immediate orientation
  • Compressed Prerequisites — bullet list instead of numbered bold items
  • Consolidated "Strategic Purpose" into Point Tags — eliminated the separate numbered section for categories, merged into the per-witness template as Purpose
  • Trimmed Expert section — removed the fifth bullet (reliance on incomplete data) which overlaps with "unsupported factual assumptions"
  • Renamed Guidelines → Rules — shorter, more directive
  • Removed redundant prose — e.g., "For each witness, organize by theme (not chronology)" moved to Quick Start instead of appearing as a standalone instruction
  • ~90 lines → ~80 lines with better information density and clearer progressive structure
저장한 Skills