MCPProxy
Local-first MCP proxy with BM25 tool discovery, security scanning, quarantine & ~99% token savings
Description from the maker (original English)
Honeypot probe data: IP reputation, scanners, CVE probing, TLS and SSH fingerprints.
These three are never merged into one score: opinions, measurements and popularity are different things.
We have not written a plain-language summary for this one yet. Below are the facts we read from the registry.
Pick the AI app you use, then copy the settings.
AuthorizationValues marked secret are keys to your account. Never share them or paste them into a chat.
You need
Which app do you use?
Pick your app above. We remember it, so next time the button says “Install in …”.
This address needs a key in a request header, which the add-connector menu cannot set. Use Claude Code, Cursor or VS Code instead.
Run this in your terminal
claude mcp add --transport http honeylabs https://mcp.honeylabs.net/mcp --header 'Authorization: your-value'Fill in the empty values before running it.
Paste into .cursor/mcp.json in your project, or ~/.cursor/mcp.json
{
"mcpServers": {
"honeylabs": {
"url": "https://mcp.honeylabs.net/mcp",
"headers": {
"Authorization": "your-value"
}
}
}
}Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.
Paste into .vscode/mcp.json in your project
{
"servers": {
"honeylabs": {
"type": "http",
"url": "https://mcp.honeylabs.net/mcp",
"headers": {
"Authorization": "your-value"
}
}
}
}Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.
This address needs a key in a request header, which the add-connector menu cannot set. Use Claude Code, Cursor or VS Code instead.
Your choice stays on this device in a cookie; no account needed.
Authorization secret required HTTP headerThese settings are generated from the maker’s registry entry. If the README says otherwise, follow the README.
Tool names and descriptions exactly as the server reported them on Nov 7, 2026. The maker wrote them, not us; we show them as plain text.
search_events_toolIndividual honeypot events with every field, newest first, filtered by source, network, port, protocol or fingerprint. Use when the user wants to see the records themselves: "show me the events from this IP", "what hit port 443 yesterday", "every request carrying this JA4", "all
top_attackers_toolRanked leaderboard over a time window: the top source addresses, networks, countries, destination ports, user agents, JA4 fingerprints, URL paths, reverse DNS domains or CVEs by event count. Use for "who is attacking the most", "top countries hitting port 22", "most targeted port
ioc_lookup_toolEverything the sensors hold on one IP address, a CIDR range, several ranges at once, or a domain, with our verdict and the reasons for it. Start here for any question about a specific indicator. Use for "is this IP malicious", "is this a known scanner", "have you seen this addres
cve_lookup_toolWho is probing one named CVE, and how: counts, top probing addresses, networks, countries, fingerprints, sample request paths and a daily timeline. Use whenever the user names a CVE: "is CVE-2024-4577 being exploited in the wild", "who scans for CVE-2023-1389", "has anyone probed
payload_search_toolLiteral substring search, ignoring case, over captured request text: URL path, request body, request headers and event summary. Use for "find requests to /wp-admin", "which requests carried this user agent string", "who sent '../../etc/passwd' last week". It matches text that app
attack_timeline_toolEvent counts over time, in hourly or daily buckets, optionally filtered by protocol, country or destination port. Use for "show the trend this week", "was there a spike on port 22", "how has SSH scanning changed over 30 days". Do not use it to find who caused a spike (follow up w
asn_enrich_toolProfile of one autonomous system (a hosting provider, ISP or company network) over a time window: total events, unique addresses, top targeted ports, source countries, top user agents and the organisation name. Use for "tell me about AS202425", "what is this hosting provider doin
fingerprint_search_toolEvents and sources that carried one client fingerprint in a time window. Use for "have you seen this JA4", "which IPs share this TLS fingerprint", "find every scanner with this SSH client fingerprint". For how widespread a fingerprint is and whether it is one operation or a commo
fingerprint_similar_toolHTTP request shapes within a few headers of one Akin fingerprint, with what those clients request and call themselves, and the family the fingerprint belongs to. Use when an unfamiliar request shape turns up in your own web server, WAF or proxy logs and you want to know whether i
fingerprint_population_toolThe population behind one client fingerprint over the full retained window: how many source addresses carry it, across how many networks and countries, the ports they hit, the top networks, a sample of addresses, and a read on its shape. Use for "is this JA4 one botnet or a commo
udp_activity_toolUDP datagrams sent to the sensors: DNS questions, QUIC client fingerprints and other UDP probes, with the ports, networks and countries involved. Use for "who is sending DNS ANY queries", "is anyone querying version.bind", "have you seen this QUIC JA4", "which UDP ports are probe
interactive_credentials_toolPlaintext logins captured by the interactive sensors (sensors that answer as a real service and hold the session): username, password, the service they were tried on, the outcome and the source address, newest first, plus the pairs tried most and a per-service breakdown. Use for
interactive_files_toolFiles clients handed to the interactive sensors: uploads, mail attachments, published messages and other artifacts, by SHA-256, with name, size, MIME type, the sessions that carried them and the VirusTotal verdict when one is known. Use for "have you seen this hash", "what files
interactive_sessions_toolSessions one address or range held with the interactive sensors: when, which port and protocol, the service the sensor answered as, the session verdict (scan, credential or exploit), labels, turns, bytes, duration, CVE ids proven by the payload, path-only CVE candidates and the d
We check information anyone can see; we have not read the code. Checks that do not apply are left out of the count. · Checked Nov 8, 2026
The registry name was verified against the domain honeylabs.net (the registry checks domain ownership)
We have not read this repo from GitHub yet
We have not read this repo from GitHub yet
Responded in 72 ms (HTTP 200)
No sign-in needed
Read 14 tool descriptions and found no hidden instructions to the AI
Compared with every active server in this category in the registry. Stars count only when the publisher owns the repo.
It passed 3 of 3 applicable basic checks. We have not read the code. Before adding a key or token, see what access the server asks for and use a key with only the permissions you need.
You need to fill in: Authorization
It runs in the maker’s cloud. Connect over the internet with nothing to install, but what you send passes through the maker’s servers.
The GitHub repo last changed –, and the latest version (1.2.0) reached the registry 9 days ago.
Apps that support MCP, such as Claude Desktop, Claude Code, Cursor and VS Code. It connects by URL, so apps that accept remote MCP servers work too, such as ChatGPT in developer mode.
In the same category: MCPProxy, ConnectSecure MCP, EchelonGraph CVE & Exposure
Servers in the same category from other makers, sorted by stars.
Reading a card:★ GitHub starsiThree pillars: reviews · checks · usageiUpdated = latest version published to the registryi
Local-first MCP proxy with BM25 tool discovery, security scanning, quarantine & ~99% token savings
Vulnerability management across all ConnectSecure clients: exposure ranking, CVE impact, blind spots
CVE, KEV, EPSS, SBOM and advisory lookups; per-CVE exposure from Shodan data (© Shodan). Keyless.
Copied from the maker’s repo as written. PopNx MCP has not reviewed or endorsed it. HTML, images and scripts are removed. Links lead off this site.
Honeypot threat intelligence as MCP tools. Query 90 days of probe data from our honeypot sensor network: IP reputation, scanner classification, CVE probing trends, TLS/HTTP/SSH fingerprints (JA4, JA3, Akin, HASSH), mTLS client certificates, Community ID flow hashes, and attack timelines. Use it straight from Claude, Cursor, Gemini, Cline, or any other Model Context Protocol client.
claude mcp add honeylabs \
--transport http \ https://mcp.honeylabs.net/mcp \ --header "Authorization: Bearer "
Get a key at https://honeylabs.net/dashboard?src=github (magic-link sign-in, no password).
Add to your MCP config:
{
"mcpServers": { "honeylabs": { "url": "https://mcp.honeylabs.net/mcp", "headers": { "Authorization": "Bearer " } } } }
Same JSON config as Claude Desktop / Cursor. Install via the MCP Marketplace listing or paste the config block above into your settings.
gemini /mcp add honeylabs https://mcp.honeylabs.net/mcp
gemini /mcp auth honeylabs # OAuth flow, no static key
OAuth 2.1 with PKCE + DCR is supported at /oauth/authorize. Any MCP client that speaks standard OAuth (Gemini, MCP Inspector, Smithery, Cline's OAuth flow) works out of the box.
| Tool | What it answers |
|---|---|
ioc_lookup | Is this IP / domain known to be probing? Returns our verdict, whether it is a recognised benign scanner, the CVEs it probed, plus ports / paths / fingerprints. |
top_attackers | Ranked leaderboard of source IPs, ASNs, countries, ports, user-agents, or CVEs (by='cve') over a time window. |
search_events | Raw honeypot events matching filters (IP, ASN, country, dest_port, protocol, http_method, ja4/ja3/akin, community_id, has_client_cert). |
attack_timeline | Hourly / daily attack volume over a window, with protocol / country / port filters. |
asn_enrich | Full profile for an ASN: total events, unique IPs, top ports, source countries, user-agents, org name. |
fingerprint_search | Search by TLS JA4 / JA3, HTTP Akin or SSH HASSH fingerprint to find shared infrastructure. |
fingerprint_similar | Request shapes within a few headers of an Akin HTTP fingerprint, what those clients ask for, and the family the token belongs to. |
cve_lookup | Who is probing a named CVE: severity, KEV status, top probing IPs with ASN and scanner tag, fingerprints, sample paths, daily timeline. |
payload_search | Substring search across request paths, bodies and headers. Never a CVE id. |
fingerprint_population | How many addresses and networks carry one fingerprint, and whether it looks like one operation. |
udp_activity | DNS questions, QUIC client fingerprints and other UDP probes. UDP sources can be forged. |
interactive_credentials | Logins tried against the interactive sensors, per service. |
interactive_files | Files clients handed to the interactive sensors, by SHA-256, with the VirusTotal verdict. |
interactive_sessions | Sessions one address or range held with the interactive sensors. |
Every tool is free to call. Plans differ by how much you can pull, not by which tools exist: each row in a response counts as one credit, and a free key gives 500 credits a day, with higher limits for heavier use. See https://honeylabs.net/docs#plans for the breakdown, or https://honeylabs.net/pricing.json for the same as JSON.
Tool names on the server carry a _tool suffix (ioc_lookup_tool). The server card is at https://mcp.honeylabs.net/mcp/server-card.
HoneyLabs runs a fleet of honeypots that get probed by the public internet all day. Every probe, meaning every connection, TLS handshake, and HTTP request, is logged with the source IP, ASN, geo, TLS/HTTP/SSH fingerprints, and full URL path. We retain the last 90 days and expose it through this MCP server, a JSON API, a public lookup web UI at /lookup/ , and CSV / STIX exports.
This is our own ground-truth record of what is actively scanning the internet right now, gathered first-hand rather than copied from a CVSS database or a third-party reputation feed.
Things to ask Claude / Cursor / Gemini once HoneyLabs is wired in:
More worked examples at https://honeylabs.net/mcp.
The honeypot fleet itself (Spip-Go) and the enrichment pipeline (Loom) are public. This repo (the MCP / API surface) is closed.