LegalLitigation

UK Witness Statement Review

Use when users say "review this witness statement", "check this statement before service", "PD 57AC", "statement of truth", "hearsay in this witness evidence", or need England and Wales witness eviden…

LLegalQuants·Legal·Apache-2.0

Library skill — the default version is maintained in GitHub; edits you make live in your own clone.

Use this skillDownload .zip
How does this work?
  • ChatGPT opens a new chat with the skill loaded. If it's too long for a link, it's copied to your clipboard — just paste.
  • Claude works the same way. To install it permanently, download the .zip and upload it under Claude → Settings → Capabilities → Skills (Pro/Team/Enterprise).
  • Copy prompt copies the skill so you can paste it into any assistant, including Grok.

uk-witness-statement-review

When to Use

This skill is a review workflow. It does not certify truth, settle admissibility, or replace solicitor/counsel sign-off.

Audience and Work Shape

Audience: England and Wales litigation lawyers, pupils/trainees, and paralegals reviewing witness evidence under supervision before service or filing.

Work shape: pattern-matched review with accretive judgment. The skill classifies paragraph types, flags support/rule issues, and prepares questions for the responsible lawyer/witness; it does not certify the evidence.

Legal Failure Modes

Access Modes

This skill works in three modes:

  1. Live source mode - use browser, web search, MCP, API, or other configured access to retrieve public rules, orders, filings, or source documents.
  2. User-supplied source mode - use the draft statement plus uploaded or pasted exhibits, pleadings, correspondence, chronology, disclosure, prior statements, rules, or orders supplied by the user.
  3. No-source mode - review the statement's internal drafting issues and produce a source-request checklist, but do not mark factual assertions as supported.

If exhibits, documents, rules, or orders cannot be retrieved or supplied, mark the relevant check source_missing or verify_current_rule. Do not verify facts or procedural compliance from model memory.

How It Works

1. Confirm context

Identify:

If the governing order or forum is unknown, flag CPR/PD points as verify_current_rule.

If the forum is a tribunal or specialist court not governed by the CPR witness statement regime, identify the applicable procedural rules and mark CPR/PD checks as not_applicable_or_verify.

2. Build a statement map

For each paragraph, classify:

Keep paragraph numbers stable. The output should let a drafter jump straight to the problem paragraph.

3. Check source support

For each factual assertion:

4. Check drafting and compliance risks

Flag:

Do not invent procedural requirements from memory. If current CPR/PD wording matters, retrieve it or mark verify_current_rule.

Confidence Bands

Escalation

Stop and route to the responsible lawyer when evidence is hearsay-critical, a source conflict affects a material fact, PD 57AC or a court order is unclear, the statement contains argument or legal conclusions that drive the case, or the user asks whether the statement is ready to sign/file.

5. Produce a review report

Recommended output:

| Paragraph | Issue | Severity | Source / Pinpoint | Rule / Source Checked | Suggested Fix | |---|---|---|---|---|

Severity values:

Then include:

Example

Review this draft witness statement against the documents and chronology. Flag unsupported paragraphs, argument, hearsay, date conflicts, and CPR/PD compliance issues. Do not rewrite the statement unless I ask.

For a compact output pattern, see examples/output.md. For paragraph classifications and witness-evidence review axes, see references/witness-review-playbook.md.

Limitations