decision-record-verification
Cross-references a judge's decision, court opinion, or final order against one or more hearing/trial/deposition transcripts to verify whether the record supports each finding. Produces strict dual-citation blocks with document names, page/line references, and verbatim quotes from both sources. Use when preparing appellate review, post-trial motions, record discrepancy audits, or transcript-based fact checks of judicial findings.
Best use case
decision-record-verification is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
Cross-references a judge's decision, court opinion, or final order against one or more hearing/trial/deposition transcripts to verify whether the record supports each finding. Produces strict dual-citation blocks with document names, page/line references, and verbatim quotes from both sources. Use when preparing appellate review, post-trial motions, record discrepancy audits, or transcript-based fact checks of judicial findings.
Teams using decision-record-verification should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/decision-record-verification/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How decision-record-verification Compares
| Feature / Agent | decision-record-verification | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Cross-references a judge's decision, court opinion, or final order against one or more hearing/trial/deposition transcripts to verify whether the record supports each finding. Produces strict dual-citation blocks with document names, page/line references, and verbatim quotes from both sources. Use when preparing appellate review, post-trial motions, record discrepancy audits, or transcript-based fact checks of judicial findings.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Decision Record Verification Cross-reference a court decision against transcripts and surface record alignments, discrepancies, and gaps with exact dual-source citations. ## Prerequisites - **Decision document** — opinion/order/judgment to test against the record - **Transcript set** — one or more sessions (trial, hearing, or deposition) with page/line numbering when available - **Session labels** — clear names for each transcript file (for example, `Day 1 AM`, `Day 1 PM`) - **Scope confirmation** — identify whether review is full-document or limited to specific findings/issues If required documents are missing or unlabeled, request them before analysis. ## Quick Start 1. Read the decision in full and extract each factual finding, credibility determination, and testimony-dependent legal conclusion. 2. Convert each finding into a verification claim. 3. Review transcripts in session order and capture support/conflict passages for each claim. 4. Write one citation block per match using exact quotes and pinpoint citations. 5. Finish with a findings summary: alignments, discrepancies, and record gaps. ## Workflow ### 1. Anchor the Decision - Treat the decision as the baseline source. - Extract only claims that depend on testimony or the trial/hearing record. - Keep each claim atomic (one proposition per claim) to avoid mixed citations. ### 2. Cross-Reference Transcripts - Review each transcript session in chronological order. - Tag each matched passage as: - **`[ALIGNMENT]`** — transcript directly supports the decision's characterization - **`[DISCREPANCY]`** — transcript contradicts, materially differs from, or fails to contain what the decision attributes to the record ### 3. Build Citation Blocks Use this exact field order for every block: - **`[ALIGNMENT]` or `[DISCREPANCY]`** - **Document Name (Decision):** full decision filename/title - **Location (Decision):** `Page X, Line Y` (or paragraph/section when line numbers do not exist) - **Decision Quote:** exact verbatim excerpt - **Document Name (Transcript):** full transcript filename/session label - **Location (Transcript):** `Page X, Line Y` - **Transcript Quote:** exact verbatim excerpt - **Note:** optional 1-2 sentence explanation of why the pair is support or divergence ### 4. Summarize Findings After all citation blocks, provide: - **Primary Alignments** — highest-impact ways the record supports the decision - **Primary Discrepancies** — highest-impact ways the record diverges from the decision - **Record Gaps** — decision findings with no located supporting/opposing transcript passage ## Output Order 1. **Header block** listing all reviewed documents and transcript session labels 2. **Citation blocks** grouped by session, in session order 3. Within each session group: list **`[ALIGNMENT]`** blocks first, then **`[DISCREPANCY]`** blocks 4. **Large-corpus override:** if the corpus is very large or the user requests priority triage, list **`[DISCREPANCY]`** blocks first, then **`[ALIGNMENT]`** blocks 5. **Summary of Findings** ## Quality Controls - Quote verbatim in both `Decision Quote` and `Transcript Quote`; do not paraphrase those fields. - Never invent page/line references. - If pinpoint location cannot be determined, write: - `Page/Line unavailable — approximate location: <section or paragraph description>` - Do not treat similar language as equivalent to an exact statement. - If a decision paraphrases testimony, flag that in `Note`. - Flag sessions with zero relevant passages; absence can be material. - Mark uncertain citations or interpretations with `[VERIFY]`. ## Edge Cases - **Multi-session transcripts**: treat each file as a distinct source and keep session label in every block. - **Incomplete/redacted transcripts**: record the limitation under `Record Gaps`. - **Privileged or sealed material**: do not reproduce protected excerpts; flag and advise handling under governing order. - **Non-U.S. records**: adapt pinpoint format to the jurisdiction's transcript citation convention.
Related Skills
managing-medical-records-compliance
Evaluates medical records practices against retention, access, and amendment requirements. Use when auditing medical records, managing record retention, or processing amendment requests.
designing-clinical-decision-support
Creates CDS rule specifications with trigger conditions, evidence base, and alert fatigue mitigation. Use when designing CDS alerts, specifying clinical rules, or optimizing alert systems.
record-designation
Drafts a Record Designation on Appeal identifying trial court materials for transmission to the appellate court. Covers appellant designations and appellee counter-designations with transcript requests, exhibit lists, cost allocation, and compliance guidance. Use when filing a designation of record, counter-designation, transcript designation, or preparing the appellate record after a notice of appeal.
medical-record-chronology
Creates verified, Bates-cited chronological summaries of medical records for litigation. Extracts treatment timelines, providers, diagnoses, procedures, medications, and causation evidence with every clinical fact traceable to a specific Bates range. Identifies gaps, inconsistencies, pre-existing conditions, and records needing follow-up. Flags unverifiable claims rather than dropping them. Use this skill whenever the user asks for a medical chronology, medical summary, treatment timeline, or records review in a personal injury, medical malpractice, workers comp, or disability case, and any time medical records are produced for analysis, even if the user does not say the word 'chronology'.
expert-medical-record-omissions
Analyzes expert witness reports, depositions, and testimony to identify gaps in medical record review, map conflicts to the full record set, and produce impeachment and admissibility analysis with pinpoint citations. Use when reviewing expert reports for missing medical records, selective review patterns, Daubert/Frye challenges, cross-examination prep, motions to exclude or limit, or rebuttal expert focus in personal injury or medical malpractice litigation.
discovery-verification-audit
Produces an attorney-grade audit memorandum assessing whether written discovery responses are legally binding and timely. Use this skill when the user mentions verification audit, discovery verification, proof of service review, signature authority, perjury clause compliance, deadline computation, mailbox rule extensions, service defects, waiver risk, or motion-to-compel deadline analysis. Also trigger when the user references FRCP 33(b)(3) verification, CCP 2015.5, 28 U.S.C. 1746, attorney vs. party verification, or asks for help checking whether discovery responses are properly executed. Even if the user just says "check these responses for defects" or "is this verification valid," use this skill.
ca-discovery-verification
Drafts California Superior Court discovery verification pages under CCP §§ 2030.250 and 2031.250 with proper CCP § 2015.5 perjury declarations. Trigger when the user needs a verification page for interrogatory or inspection/RPD responses, mentions California discovery verification, party vs. attorney signing, information-and-belief clauses, or fixing unverified responses. CA Superior Court only — not federal.
bates-citation-verification
Citation format and verification protocol for legal documents that draw on Bates-stamped source records. Defines the bracket citation format, multi-source citation rules, quotation handling, the verification pass that runs before output, and the [UNVERIFIED] flag protocol. Use this skill any time another skill produces a legal work product that cites Bates-stamped pages, including medical chronologies, deposition summaries, demand letters, case briefs, and discovery responses. Also use any time the user asks how to format a citation or verify that a claim is supported by the source.
authority-verification
Verifies legal citations and retrieves source material using the case.dev Legal Research API and CLI. Use when validating a citation, confirming a case exists, pulling source text, finding related authorities, extracting citations from a brief or opinion, or building a filing-ready authority check without branded citator terminology.
modeling-capital-budgeting-decisions
Builds NPV, IRR, and payback models for capital investment decisions with hurdle rate calibration and risk adjustment. Use when evaluating capital investments, comparing project returns, or building capital allocation frameworks.
managing-follow-on-investment-decisions
Evaluates pro-rata and follow-on investment decisions with portfolio construction, signaling effects, and reserve management. Use when making follow-on decisions, managing reserves, or evaluating pro-rata rights.
managing-bridge-financing-decisions
Evaluates bridge round scenarios with insider dynamics, signal risk, and term structuring for existing portfolio companies. Use when considering bridge financing, structuring insider rounds, or analyzing pay-to-play provisions.