H1B Sponsor MCP
Read-only MCP server for querying H-1B sponsoring employers from the USCIS dataset, providing tools to search employers, get year-by-year approvals, top sponsors, and trends across FY2009-2026.
README
H1B Sponsor MCP
Read-only MCP server for querying H-1B sponsoring employers from the USCIS Employer Information dataset.
Repository: https://github.com/ujjwalredd/H1B-Sponsor-MCP.git

This server lets MCP clients answer questions like:
- "Does Stripe appear in the H-1B employer data?"
- "Which companies had the most H-1B approvals in Texas in FY2024?"
- "How did manufacturing sponsorship change from FY2009 to FY2026?"
The bundled cleaned dataset covers fiscal years 2009-2026 with 1,055,650 employer-year/location records and 392,843 distinct employer name strings. FY2026 is a partial year.
What It Provides
| Tool | Purpose |
|---|---|
search_employers |
Find employers by partial name, with lifetime totals |
employer_profile |
Get year-by-year approvals, denials, locations, and sectors for matching employers |
top_sponsors |
Rank sponsors by approvals, denials, petitions, approval rate, new employment approvals, or continuation approvals |
yearly_trends |
Return per-year totals from FY2009-FY2026, optionally filtered by state or NAICS sector |
industry_breakdown |
Summarize sponsorship by NAICS sector |
state_breakdown |
Summarize sponsorship by US state or territory |
dataset_info |
Return dataset coverage, row counts, and caveats |
Quick Start
uvx can fetch and run the server directly from
GitHub:
uvx --from git+https://github.com/ujjwalredd/H1B-Sponsor-MCP.git h1b-sponsor-mcp
After the package is published to PyPI, this shorter form should also work:
uvx h1b-sponsor-mcp
Install uv if needed:
curl -LsSf https://astral.sh/uv/install.sh | sh
Claude Code
claude mcp add h1b-sponsors -- uvx --from git+https://github.com/ujjwalredd/H1B-Sponsor-MCP.git h1b-sponsor-mcp
Claude Desktop
Add this server to claude_desktop_config.json:
{
"mcpServers": {
"h1b-sponsors": {
"command": "uvx",
"args": [
"--from",
"git+https://github.com/ujjwalredd/H1B-Sponsor-MCP.git",
"h1b-sponsor-mcp"
]
}
}
}
Manual Install
git clone https://github.com/ujjwalredd/H1B-Sponsor-MCP.git
cd H1B-Sponsor-MCP
python -m pip install -e .
h1b-sponsor-mcp
The cleaned parquet is included in the package. To use a different cleaned dataset file:
export H1B_DATA_PATH=/path/to/h1b_employers_clean.parquet
H1B_DATA_PATH must point to a .parquet file.
Data
- Source: USCIS H-1B Employer Data Hub, Employer Information exports.
- Coverage: fiscal years 2009-2026.
- Packaged file:
src/h1b_mcp/data/h1b_employers_clean.parquet. - Details: see DATA.md.
Recommended source citation:
U.S. Citizenship and Immigration Services (USCIS). H-1B Employer Data Hub, Employer Information exports. https://www.uscis.gov/tools/reports-and-studies/h-1b-employer-data-hub
This project is independent and is not affiliated with, sponsored by, or endorsed by USCIS, DHS, or the U.S. Government. The source data comes from a U.S. federal government publication; U.S. Government works are generally not copyright-protected in the United States under 17 U.S.C. Section 105. This is a practical attribution note, not legal advice.
Important caveats:
- FY2026 is partial.
- Counts are petitions, not unique workers.
- Employer names are not entity-resolved; the same company may appear under multiple spellings or legal names.
approval_rateisapprovals / (approvals + denials).- This project is an information tool, not legal or immigration advice.
Security Model
This server is intentionally narrow:
- stdio transport by default; no network listener is opened by this package.
- Read-only data access; no mutating MCP tools.
- No SQL,
eval,exec, pickle loading, or shell execution. - User search text is escaped before regex matching.
- Years, states, NAICS sectors, sort metrics, and limits are validated before the data layer runs.
- Results are hard-capped to protect client context windows.
- Errors returned to clients are sanitized; stack traces stay in stderr logs.
See SECURITY.md for the full policy and vulnerability reporting instructions.
Development
python -m pip install -e ".[dev]"
pytest
ruff check src tests
For quick local test runs without installing the editable package:
PYTHONPATH=src pytest
Run the MCP inspector:
mcp dev src/h1b_mcp/server.py
Open Source
- Contributions: CONTRIBUTING.md
- Code of conduct: CODE_OF_CONDUCT.md
- Support: SUPPORT.md
- Security: SECURITY.md
- Notice and attribution: NOTICE.md
- License: MIT
The source code is MIT licensed. Retain USCIS attribution when redistributing the packaged or derived data files.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.