Claude Code skills for immigration attorneys and self-petitioning candidates.
Draft O-1A, EB-1A, and EB-2 NIW petitions with AI — every claim backed by evidence.
A set of Claude Code skills that handle the most time-consuming parts of immigration petition work. Built for two audiences:
For immigration attorneys:
- Organize 100+ client documents into a structured evidence index
- Optionally run a pre-draft check on which criteria or NIW prongs look supportable from the index (skip if you prefer to go straight to drafting)
- Optionally build a normalized publication and citation table for research-heavy O-1A, EB-1A, or NIW cases
- Draft petition letters with proper legal structure, exhibit references, and evidence enrichment
- Audit drafts for factual errors, unsupported claims, and legal issues before signing
- Draft expert/recommendation letters with strategic criteria assignment
For self-petitioning candidates (EB-1A, NIW):
- Understand what evidence you need and how strong your case is
- Research government sources that support your national importance argument
- Generate a first draft of your self-petition letter to review with your attorney
- Identify evidence gaps before you file
| Skill | What It Does |
|---|---|
document-summary-arrangement |
Ingests source documents (PDFs, images, scans), triages by importance, summarizes key documents, and produces a structured index arranged by evidentiary criterion |
case-strength-assessor |
Optional. Reads the document index and outputs a frank pre-draft assessment of which O-1A/EB-1A criteria or NIW prongs are strong, moderate, weak, or unsupported — use only if you want that extra pass |
publication-citation-table |
Optional. Summary table + TSV, lead picks for C5, C3 side list — from the index and optional Scholar/ORCID; tags Criterion 5 vs press (Criterion 3) and NEEDS VERIFICATION for publication-heavy cases |
o1-petition-narrative |
Drafts O-1A/O-1B petition support letters — criterion-by-criterion arguments with exhibit references and evidence enrichment |
eb1a-petition-narrative |
Drafts EB-1A green card petitions — same criteria as O-1A but stricter scrutiny, adds sustained acclaim totality argument, intent to continue work, and substantial U.S. benefit |
niw-national-importance-research |
Researches authoritative gov sources, national plans, and federal funding — produces a quotable research document matched to the petitioner's background |
niw-petition-narrative |
Drafts EB-2 NIW self-petition support letters using the Dhanasar three-prong framework |
expert-letter-drafter |
Drafts expert/reference/recommendation letters — assigns criteria to experts by credential, ensures no two letters overlap, produces ready-to-customize drafts |
petition-audit |
Audits any petition letter — verifies every factual claim against source documents, flags legal errors, checks internal consistency, produces severity-rated report |
rfe-response-drafter |
Analyzes a USCIS RFE, diagnoses what went wrong in the original petition, maps each deficiency to a fix, and drafts the point-by-point response with new evidence strategy |
The knowledge/ directory contains best practices distilled from real cases and USCIS adjudication patterns. Skills read these files before drafting.
knowledge/
├── overview-o1a-eb1a.md # O-1A/EB-1A standard, Kazarian framework
├── overview-niw.md # NIW standard, Dhanasar framework
├── criteria/ # Best practices per O-1A/EB-1A criterion
│ ├── 01-awards.md
│ ├── 02-membership.md
│ ├── 03-published-material.md
│ ├── 04-original-contributions.md
│ ├── 05-scholarly-articles.md
│ ├── 06-critical-employment.md
│ ├── 07-high-remuneration.md
│ └── 08-judging.md
├── prongs/ # Best practices per NIW Dhanasar prong
│ ├── 01-substantial-merit.md
│ ├── 02-well-positioned.md
│ └── 03-national-interest-balance.md
├── evidence-hierarchy.md # How USCIS weighs evidence (Tier 1–4)
├── argument-patterns.md # Reusable argument structures
└── uscis-policy-alerts.md # Key citations (PA-2022-03, Kazarian, Dhanasar)
Source documents
↓
document-summary-arrangement → document index (arranged by criterion)
↓
knowledge/ (read best practices)
↓
o1-petition-narrative → O-1A petition package ─┐
eb1a-petition-narrative → EB-1A petition package ─┤
niw-national-importance-research → research document │
→ niw-petition-narrative → NIW petition package ──┤
expert-letter-drafter → expert letter drafts │
↓
petition-audit → audit report
↓
Attorney reviews & signs
Optional, any time after the index exists (often before or during narrative drafting): case-strength-assessor — pre-draft criterion/prong strength; publication-citation-table — master publication/citation table for research-heavy petitions. Neither is required; skip if you do not need them.
Here's what the skills actually produce — from a real test run on the Judging criterion:
Every entity is researched with facts traced to source documents and verified web URLs:
Web sources are compiled with exact quotes, retrieval dates, and attorney action checkboxes:
The final draft uses ONLY sourced facts — every number has a citation:
Each petition package is a single file containing:
- The full support letter draft
- Source registry (every fact traced to a document or URL)
- Web sources the attorney needs to print for filing
- Evidence gaps and strength assessment
- Install Claude Code
- Clone this repo
- Install dependencies:
./document-summary-arrangement/scripts/setup.sh
- Point Claude at your client's documents:
/document-summary-arrangement - Optionally assess case strength from the index before drafting:
For research-heavy clients, optionally normalize publications and citations:
/case-strength-assessor/publication-citation-table - Draft the petition:
/o1-petition-narrative # O-1A temporary visa /eb1a-petition-narrative # EB-1A green card /niw-petition-narrative # EB-2 NIW - Draft expert letters:
/expert-letter-drafter - Audit before signing:
/petition-audit
- Install Claude Code
- Clone this repo
- Gather your documents (awards, memberships, publications, employment records, funding docs)
- Index your documents:
/document-summary-arrangement - Optionally see which criteria or prongs look strong or weak before drafting:
If you have many papers, optionally build a publication/citation table from the index:
/case-strength-assessor/publication-citation-table - For NIW — research your national importance argument:
/niw-national-importance-research - Draft your self-petition:
/eb1a-petition-narrative # EB-1A self-petition /niw-petition-narrative # NIW self-petition - Audit the draft:
/petition-audit - Take the draft to an immigration attorney for review before filing.
Helper scripts for document processing in document-summary-arrangement/scripts/:
| Script | What It Does |
|---|---|
setup.sh |
Installs dependencies (poppler for PDF tools) |
scan-collection.sh |
Scans a directory — reports file types, sizes, page counts, batch plan |
pdf-to-text.sh |
Extracts text from PDF (auto-detects scanned vs. text) |
pdf-to-images.sh |
Converts PDF pages to PNG for multimodal reading |
docx-to-text.sh |
Extracts text from .docx (macOS/Linux) |
- Immigration attorneys drafting O-1A, EB-1A, and NIW petitions who want structured, evidence-backed first drafts
- Self-petitioning candidates (EB-1A, NIW) who want to understand their case strength, organize evidence, and prepare a draft before engaging an attorney
- Legal teams that want to reduce time spent on document organization, research, and initial drafting
- Not legal advice. Every output is a draft. An immigration attorney must review, edit, and sign before filing.
- Not a replacement for an attorney. Self-petitioning candidates should use these tools to prepare — then take the output to a licensed immigration attorney.
- Not guaranteed to result in approval. Petition outcomes depend on the facts, the evidence, and USCIS adjudication.
The knowledge base improves with every case. After a case outcome:
- Add anonymized best practices to the relevant
criteria/orprongs/file - Never commit client names, company names, or identifying details
- Keep entries concise and actionable
For professional immigration guidance (EB-1, NIW, O-1), you can contact Neo Global.
MIT
