Legal Text Format

• You have raw statutory text (民法典, 刑法, etc.) pasted from a document or fetched by another skill, and need clean, navigable Markdown with properly bolded articles and spaced clauses.
• You are compiling a set of court decisions or Supreme Court model cases and want consistent heading levels, Chinese punctuation, half-width numerals, and all social-media filler stripped out.
• You are building a legal knowledge base or study materials and need each document archived with timestamped filenames and a metadata header linking back to the source.
• Another skill (e.g. wechat-article-fetch) has retrieved a WeChat legal article and needs to hand it off for formatting — this skill accepts that file path directly.

• You need to retrieve legal content from a URL or WeChat link — this skill cannot fetch web content; use a dedicated scraping skill first.
• Your input is non-Chinese legal text; the formatting rules (Chinese punctuation, 章/条 structure, half-width numeral conversion) are designed specifically for Simplified Chinese legal documents.

The skill uses an LLM prompt to classify incoming text as either a legal statute or a legal case before applying the appropriate formatting ruleset. This means you don't need to specify the type — the skill adapts automatically based on structural signals like chapter headings, article numbers, or case docket patterns.

For statutory text, chapters receive ## headings, each article (第X条) is bolded, blank lines are added between articles, and any trailing line-breaks without a closing punctuation mark are cleaned up. All substantive content is preserved verbatim — no clause is ever truncated.

For case documents, each case title or sequence number gets a ## heading, sub-sections (facts, ruling, significance) get ### headings, English punctuation is replaced with Chinese equivalents, and numbers are normalized to half-width. Introductory text, author bios, footers, QR-code promotions, and recommended-article blocks are automatically stripped.

Every run saves both the raw input and the formatted output into a timestamped subdirectory under archive/. Files follow a predictable naming convention (YYYYMMDD_主题_raw.md / _formatted.md), making it easy to audit changes or recover the original text. An optional meta.json captures run metadata.

When wechat-article-fetch (or another fetch skill) detects legal content and this skill is installed, it can pass a file path directly as input. The skill reads the saved Markdown, runs the full format pipeline, and writes results to its own archive/ directory — no manual handoff required.

Formatting changes are strictly cosmetic: punctuation normalization, heading levels, and whitespace. Case numbers, court names, party names, article text, and all substantive legal content are preserved exactly as received. The original file is always kept as a reference.

A user copies several chapters of the 民法典 into the chat. The skill detects statute structure, wraps each chapter in ##, bolds every 第X条, normalizes spacing, and saves raw.md plus formatted.md to a timestamped archive folder — ready to import into a knowledge base.

A legal researcher fetches a WeChat article containing 10 Supreme Court model cases. The skill strips the article intro, author bio, and trailing promotional content, then applies consistent ##/### heading hierarchy and Chinese punctuation across all 10 cases, producing a clean reference document.

A law student regularly processes judicial decisions for study. Each formatted file lands in its own dated archive directory with a metadata header that includes the source URL (when provided by a fetch skill), processing timestamp, and document type — making the library easy to search and cite.

With wechat-article-fetch installed, a user provides a WeChat public-account link. The fetch skill retrieves the article and automatically triggers legal-text-format, which formats and archives the result without any additional user input — the entire pipeline runs as a single request.