
Relayshield Firewall
io.github.nzdsf2-gifPublished by GitHub account nzdsf2-gifv0.3.0Updated Nov 22, 2026
RelayShield MCP Proxy Firewall. Screens every MCP tool call. 50 free calls, then $49/month.
Overview
RelayShield exposes threat-intelligence checks — breach, SIM swap, lookalike domains, malware scans and credential exposure — as MCP tools for an assistant.
- What you can do
- Look up email breach records, SIM swap and eSIM status, lookalike domains, and OAuth app exposure
- Check infostealer logs, stolen sessions, non-human identity credentials, and secrets in public GitHub repos
- Scan URLs and files across many engines, then poll for the verdict
- Assess MCP server registry risk and vendor supply-chain exposure
- Check EVM wallet on-chain risk
- When to use it
- Useful when an assistant needs to vet an email, phone number, domain, URL, file, wallet address or MCP server for fraud, phishing or credential exposure, or when auditing the authority an agent or vendor holds.
- Requirements
- Remote endpoint at tnt-demo.proxy.relayshield.net with OAuth required. The README also documents a local Python package (pip or uvx) configured with RELAYSHIELD_API_URL plus either RELAYSHIELD_API_KEY (subscription) or RELAYSHIELD_X_PAYMENT (pay-as-you-go).
Installation
In SourceWeft
- Open Relayshield Firewall in the dashboard and add it to a workspace.
- Connect the required credentials. They are stored privately and never shown on public pages.
- Enable the server for the chats that should use its tools.
Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.
Use with an agent
Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.
Read https://sourceweft.com/mcp/io.github.nzdsf2-gif%2Frelayshield-firewall/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.Other MCP clients
Add this to your client's mcpServers config. The client completes OAuth sign-in on first connect.
{
"mcpServers": {
"relayshield-firewall": {
"type": "http",
"url": "https://tnt-demo.proxy.relayshield.net/mcp"
}
}
}README
relayshield-mcp
mcp-name: io.github.nzdsf2-gif/relayshield-mcp
RelayShield security intelligence as an MCP server — plug breach detection, SIM swap detection, domain lookalike monitoring, OAuth supply chain watchlist, and URL/file scanning directly into Claude and any MCP-compatible AI agent.
Tools
check_oauth_watchlist, check_supply_chain, check_session_risk, check_nhi_exposure, and
check_secret_scan cover related ground — connected-app, session, and machine-credential exposure
for an identity or its supply chain — and are a natural set to use together when vetting an agent's
current authority, not just a login.
Access modes
Subscription — API key from api.relayshield.net/developers. All 15 tools available. Free tier: 100 calls/month. Paid tiers from $29/month.
Pay-as-you-go — No API key needed. Pay per check in USDC on Base (x402 protocol). Set RELAYSHIELD_X_PAYMENT with your payment proof. All 15 tools available ($0.05–$0.50/check, check_scan_result free). Call a tool with no payment set to receive pricing and payment instructions.
Discovery — Set neither key nor payment. Tool calls return payment requirements and a subscription link.
Install
Or run without installing:
Configure Claude Desktop
Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
Subscription (RelayShield API key):
Pay-as-you-go (x402 USDC on Base):
Quit and relaunch Claude Desktop after editing.
Configure Claude Code (CLI)
Free checks server (no key, no signup)
relayshield-free-mcp ships in the same package and exposes RelayShield's free,
keyless threat-intel checks as four MCP tools. It is the funnel into the paid
server above: every summary links the deeper paid checks where they apply.
Claude Desktop — add to claude_desktop_config.json:
ChatGPT — ChatGPT connects to MCP servers through Developer Mode
connectors, which require a remote HTTP endpoint; it does not launch a local
stdio server. A hosted RelayShield MCP endpoint is planned — until then, use
Claude Desktop, Claude Code, Cursor, or any other stdio-capable MCP client.
Set RS_SOURCE_TAG=mcp-free-chatgpt once the hosted endpoint is live so usage
is attributed to the ChatGPT channel. check_breach stays disabled until a
partner key is set via RELAYSHIELD_API_KEY.
Partner keys — distributors (Claude, ChatGPT, registries) each get their own
RelayShield partner key with a per-key daily call cap, set via
RELAYSHIELD_API_KEY. Separate keys per channel mean one channel's abuse never
takes down the others, and RS_SOURCE_TAG attributes usage per channel. The
server runs fine without any key — link, wallet, and email checks stay keyless
on the free tier.
Usage examples
Once configured, ask Claude:
For URL and file scans, Claude automatically polls check_scan_result every 5 seconds until the verdict is ready.
Environment variables
Set RELAYSHIELD_API_KEY or RELAYSHIELD_X_PAYMENT — not both. API key takes priority if both are set.
Links
- Landing page
- Get an API key
- GitHub
- Telegram bot / miniApp — free scam checks in chat
- WhatsApp scam-check bot — message a link, wallet, or email
Source: README.md at commit ca6400c
Tools
0Version history
1- v0.3.0LatestNov 22, 2026
Score
75% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
- Reached in the last 30 daysRequired25 pts
- Auth declaredRequired10 pts
- Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
- DescriptionRequired15 pts
- Logo10 pts
- Documentation15 pts
- License10 pts
Declare a license in the repository or the registry entry.