Log inSubmit server
Whisper logo

Whisper

Verifies Whisper agent identities, looks up RDAP records, assesses indicators, and provides agent egress over MCP.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install Whisper

Pick your client, then add it in one click or paste the config.

STDIONode.js
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "whisper": {
      "command": "npx",
      "args": [
        "-y",
        "@whisper-security/whisper-mcp"
      ]
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

Automated check passed— started and listed 5 tools correctly (1d ago).

Overview

The whisper MCP server gives agents access to Whisper identity verification, RDAP lookups, security-graph analysis, and agent networking. It can run locally through the `whisper mcp` command or remotely at `https://whisper.online/mcp` over streamable HTTP. Keyless access exposes selected verification and graph tools, while an API key adds agent provisioning, governance, and submission capabilities. Reach for it when an agent needs a routable IPv6 identity, verified egress, or Whisper security data rather than speech transcription.

Use cases

  • Verify an agent’s IPv6 identity before allowing communication
  • Look up public RDAP records for routable IPv6 addresses
  • Assess domains, IPs, ASNs, or file hashes from an AI client
  • Route agent commands through a dedicated Whisper IPv6 egress
  • Query the Whisper security graph for evidence and relationships

Key features

  • Keyless identity verification and RDAP lookup
  • Security-graph assessment and explanations
  • Local stdio and remote streamable-HTTP MCP access
  • Routable IPv6 /128 agent identities
  • SOCKS5/HTTPS and userspace WireGuard egress
  • Scriptable CLI with JSON output and exit codes

Tools5

What Whisper lets your agent do. Expand a tool for its parameters.

Verified live~602 tokens of context

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked 1d ago

Recent health check succeeded.

7/7 checks healthy over the last 44d

What it can do

From the 5 tools this server returned when we ran it.

Handles credentials · 5Makes network requests · 1Read-only tools · 5
Network at startup5 connections
Files written at startupNone
Install scriptsNone
npm provenanceNot published
Package age27 days
Maintainers2

Observed 2026-10-21 in an isolated sandbox. A description of reach, not a security audit.

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 6d ago via OSV.dev · @whisper-security/whisper-mcp (npm)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars3Stargazers on the source repository.
npm downloads1kPackage downloads in the last 30 days.
Last commit1d agoMost recent push to the default branch.
Availability100%Our rolling endpoint + install checks that succeeded.
Install checkPassedOur sandbox started it and listed its tools.
Tools exposed5Callable tools this server registers over MCP.
Directory activity1 viewsConfig copies, upvotes, and views on AllMCPs.

How Whisper works

What whisper does

The whisper MCP server connects AI clients to Whisper’s identity and security services. It can verify whether an IPv6 address or hostname represents a real agent, retrieve the public registration record for an IPv6 /128, and assess domains, IPs, ASNs, and file hashes. It also exposes security-graph queries and explanations for verdicts.

The command-line client can additionally give an agent a routable IPv6 /128 on AS219419, configure egress from that address, and verify the resulting source address. The address functions as the agent identity and is associated with DNSSEC-signed DNS, DANE pinning, and public RDAP data.

How it works

There are two MCP connection choices. A local installation runs the same MCP service with whisper mcp; clients start it as a standard command. The README also provides an npx configuration for @whisper-security/whisper-mcp. The remote option uses https://whisper.online/mcp and streamable HTTP, so clients that accept an MCP URL can connect without installing the CLI.

The local server exposes five tools without credentials: whisper_verify, whisper_rdap, whisper_assess, whisper_identify, and whisper_explain. With a key, that server lists 45 tools, including graph access, agent registration, resolver policy, egress configuration, activity lookup, and identity revocation.

The remote endpoint has a different catalog. It lists 15 keyless tools, including graph traversal, history, asset, threat-intelligence, TLS-fingerprint, and origin-related operations, plus whisper_signup. Supplying an API key adds 22 provisioning and governance tools, including whisper_agents_* and whisper_submit. The remote endpoint does not provide the local whisper_verify or whisper_rdap tools.

Setup and configuration

Install the CLI from GitHub Releases, Homebrew, Scoop, Go, mise, or the signed Debian/Ubuntu, Fedora/RHEL, and Alpine repositories. The documented installer commands fetch a release, verify its SHA-256 checksum, and perform an additional PGP check when applicable. WHISPER_CLI_BASE can override the download source used by the installer.

For MCP use, add the local npx command and package shown in the README, or point a compatible client at the remote URL. Authentication is optional for the keyless tool sets. The CLI supports browser device login or a pasted API key, and whisper login stores authentication for later use.

Tools and capabilities

  • Verify an agent address or hostname and inspect its DANE pin.
  • Retrieve public RDAP information for an IPv6 /128.
  • Assess indicators and return supporting evidence.
  • Identify an indicator’s category, vendor, or band.
  • Explain the scoring and sources behind a verdict.
  • Query the Whisper security graph through raw Cypher or cataloged recipes.
  • Connect agent traffic through SOCKS5/HTTPS or a userspace WireGuard tunnel.
  • Run commands through the selected Whisper egress and check verification with whisper ip.

The whisper ip command returns exit code 0 only when the observed egress address is in 2a04:2a01::/32 and matches the selected agent’s /128. Commands can request JSON output for scripting.

Limitations and notes

The local and remote MCP services are not equivalent: they expose different tools and have different keyless behavior. Remote access does not currently include the local identity-verification and RDAP tools. The keyed whisper_agents_verify operation verifies a domain controlled by the caller, so it is not a substitute for verifying another agent.

Creating an agent requires a name in the guided CLI flow. The CLI’s WireGuard mode is described as a userspace tunnel that does not require root; the default connection mode uses Tier-1.5 SOCKS5/HTTPS.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Whisper

whisper is an MCP server that connects AI clients to Whisper identity, RDAP, and security-graph services. Its main tools verify agent addresses, retrieve IPv6 registration records, assess indicators, identify indicator types, and explain verdicts. It runs locally through the CLI or remotely over streamable HTTP.

Install it through the documented `npx` configuration, or install the Whisper CLI and run `whisper mcp`. The CLI is available through GitHub Releases, Homebrew, Scoop, Go, mise, and signed Linux repositories. The remote MCP URL requires no local installation.

Yes, whisper works with Cursor.

Yes, keyless access is available. The local server provides five tools without a key, while the remote endpoint provides 15 keyless graph and analysis tools. An API key adds provisioning, governance, and other authenticated operations.

Run the Whisper CLI’s guided flow or use its connection commands after signing in. The flow names and creates an agent, assigns a routable `/128`, connects egress, and verifies it. `whisper connect`, `whisper ip`, and `whisper run` support scripted operation.

Related MCP servers

All in Speech-to-Text All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewWhisper AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/whisper?style=directory)](https://allmcps.com/mcp/whisper)
HTML Embed
<a href="https://allmcps.com/mcp/whisper"><img src="https://allmcps.com/api/badge/whisper?style=directory" alt="Whisper on AllMCPs" /></a>