Meta Plugin

Prajwalb0208/Akoma-Ntoso-Legal-Document-Extractor

Updated October 17, 2025
Prajwalb0208/Akoma-Ntoso-Legal-Document-Extractor

meta-plugin is a Claude Code skill in the Documentation category. Legal document metadata extraction, verification, and curation command with agents for Claude.

Key points

meta-plugin
Documentation
AI-coding skill

Meta Plugin by the numbers

Data as of Jul 7, 2026 (Skillselion catalog sync)

/plugin marketplace add Prajwalb0208/Akoma-Ntoso-Legal-Document-Extractor

/plugin install meta-plugin@meta-marketplace

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

[![Listed on Skillselion](https://skillselion.com/badge/skills/Prajwalb0208/Akoma-Ntoso-Legal-Document-Extractor/meta-plugin.svg)](https://skillselion.com/skills/Prajwalb0208/Akoma-Ntoso-Legal-Document-Extractor/meta-plugin)

Last updated	October 17, 2025
Repository	Prajwalb0208/Akoma-Ntoso-Legal-Document-Extractor ↗

What it does

Legal document metadata extraction, verification, and curation command with agents for Claude.

README.md

Akoma Ntoso All-in-One Legal Structure Extractor

This project provides a robust, hierarchical extraction pipeline for legal documents (PDF and DOCX) into fully standards-compliant Akoma Ntoso XML—including metadata, structure, and deep headings (meta, parts, chapters, articles, acts, sections, subsections, lists, and more).

Features

🏛️ Universal Extraction: Captures complete metadata (<meta>) and section structure (Parts, Chapters, Articles, Sections, Subsections, Lists) in nested Akoma Ntoso format.
🗂️ Hierarchical Output: Preserves all levels (or combinations) in the source—works with acts that have only sections, or have parts/chapters/articles, or any legal hierarchical depth.
🔍 Model-Driven Agents: Generator, Verifier (against the source), Curator (final XML), and Playbook Guardian agents for persistent learning and error avoidance.
📚 Playbook Learning: Tracks extraction/validation issues in a persistent JSON log—agents reference this on every run to continuously reduce mistakes.
🧩 Extensible: Easily adapt to new document types, languages, or regulatory structures.
⚡ Akoma Ntoso Standards: All outputs conform to Akoma Ntoso and OASIS LegalDocML schema, suitable for legal data exchange and compliance.

How It Works

Select a legal PDF/DOCX file in your workspace.
Run the /structure command:
- Extracts:
  - <meta> (metadata)
  - <body> containing <part>, <chapter>, <article>, <section>, <subsection>, <list>, etc., as present in the law
  - Only numbers and headings/titles are captured (no body/subsection prose unless required)
- Agents:
  - StructureGenerator: Extracts and maps all available hierarchy and headwords.
  - StructureVerifier: Checks structure against the PDF for order, completeness, and fidelity.
  - StructureCurator: Cleans and harmonizes all structure for Akoma Ntoso compliance.
  - StructurePlaybookGuardian: Logs issues, stores fixes, and teaches all agents—never repeat extraction, parsing, or conversion mistakes!
Download or use the generated XML for publication, search/index, or further automation.

Folder Structure

.claude/agents/: All agent definitions (generator, verifier, curator, playbook)
.claude/commands/: Command definitions (meta.md, structure.md)
.claude-plugin/: Claude plugin definitions
playbook.json: Persistent extraction/validation playbook
output/: All extracted XML results

Example Output

See the Akoma Ntoso Example for canonical structure:

<akomaNtoso>
  <act>
    <meta>...</meta>
    <body>
      <part>
        <num>Part I</num>
        <heading>...</heading>
        <chapter>...</chapter>
        <section>...</section>
      </part>
      <section>
        <num>2</num>
        <heading>...</heading>
      </section>
      <article>...</article>
      ...
    </body>
  </act>
</akomaNtoso>