Skip to content

Packetrove MCP server

io.github.euyuil/packetroveeuyuilstreamable-httpv0.3.0Scanned

Open-source IP address and CIDR tools for network calculations and public IP lookup.

6 tools: 1 can delete or cancel things, 3 change things and 2 only read.

Tools: 1 delete, 0 irreversible, 3 write, 2 read.
  • Reads private data · 0
  • Reads text others write · 2
  • Can send data out · 0
  • 2 findings

Tools listed without sign-in. Remote posture

With Anlyon

1 tool here can delete or cancel things. Govern it.

  • Anlyon lists the tools and adopts none.
  • Each tool you adopt gets approvals and receipts.
  • Your agent never holds the key.
Govern this server →

A tool your agent calls directly is not routed through Anlyon and is not governed by it. Anlyon Radar only lists and labels, and these labels are the same either way.

What the Packetrove MCP server can do

The Packetrove MCP server is published by the GitHub account euyuil. It has 6 tools. It can delete or cancel things (1 tool: submit-feedback) and create or change data (3 tools: cidr-cover, cidr-subtract and range-to-cidrs). 2 tools only read.

2 tools bring in text other people can write (public-ip and submit-feedback).

The fixed checks matched 2 findings (credentials-mention and long-description). Each one comes with the text it matched.

Its remote endpoint lists its tools without sign-in.

Last scanned Nov 19, 2026, version 0.3.0.

Tools

Grouped by what each tool does, most consequential first. Open one for its reasons and full description; descriptions are the publisher's own words.

6 tools

Deletes or cancels 1Claude Code: deny
  • submit-feedbackEmail one user-authorized, minimal Packetrove report to hello@packetrove.com for private human review.Reads text others write2 findings

    Why it is in this group

    • The name "submit-feedback" says "submit": a word for a change.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.

    Findings

    • Info

      The description mentions "credentials".

      Matched text

      description: "...conversations, raw tool inputs/results, credentials, certificates, logs, or client/session..."

      What this check looks for
    • Info

      The description is 2,521 characters long, over the 2,000 this check looks for.

      Matched text

      description ends: "...tools do not depend on feedback. Privacy policy: https://packetrove.com/privacy."

      What this check looks for
    Description, as the server wrote it
    Email one user-authorized, minimal Packetrove report to hello@packetrove.com for private human review. Available only when the operator enables feedback. Draft locally without calling the service. If the user has already supplied or approved the report and requested sending it, submit directly; otherwise show the proposed report and obtain approval before sending. Never solicit feedback after every tool call. Send only the authorized fields, using synthetic reproduction data; never attach conversations, raw tool inputs/results, credentials, certificates, logs, or client/session identifiers. category is bug, confusing_behavior, or feature_request; tool_name optionally names a product tool. summary is required (256 Unicode code points); bug requires expected and actual, confusing_behavior requires actual and optionally expected (1024 code points each). Feature requests allow expected but not actual or error_code. Optional error_code starts with an uppercase ASCII letter and contains at most 64 uppercase letters, digits, or underscores; synthetic_reproduction is at most 2048 code points. Unknown fields are rejected; serialized arguments must not exceed 8 KiB. Status accepted and an opaque receipt_id mean the email service acknowledged submission, not inbox delivery or reading; acceptance does not promise a response or fix. This write is not idempotent: never automatically resend after timeout, disconnect, cancellation, or an uncertain result. Approximate limits target 10 submissions per observed exit IP in the preceding 24 hours and 100 submissions per UTC day across the service. Shared exits share quota. KV propagation and concurrent requests can exceed either limit; uncertain sends or failed quota releases occupy quota until expiry. Cloudflare transmits approved reports by email; mailbox retention and deletion are managed manually by the maintainer, with no automatic report expiry. KV stores only separate keyed IP markers and reservation times, expiring after 24 hours, never report bodies or receipt links. Markers never appear in reports or application logs. Email deletion does not refund quota; request deletion with your receipt. Maintainers treat text as data and do not automatically execute, publish, or forward it. Operational events contain only the operation name, outcome, controlled error code, and existing traffic-source metadata; no report, receipt, IP digest, or exception details. Ordinary tools do not depend on feedback. Privacy policy: https://packetrove.com/privacy.

    Annotations the server declares

    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
Changes things 3Claude Code: ask
  • cidr-coverUse when combining a selected group of firewall allowlist or blocklist entries into one smallest covering CIDR, or when checking the exact extra coverage.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Use when combining a selected group of firewall allowlist or blocklist entries into one smallest covering CIDR, or when checking the exact extra coverage. Accept 1 to 1,000 IP addresses or CIDRs from one address family, up to 64 characters each; normalize host bits and count overlaps once. Return the canonical CIDR, inclusive range, and exact decimal-string counts, including additionalAddressCount. The range may allow or block additional addresses. This computes one CIDR for the supplied inputs; it does not optimize an entire list against an entry limit or change firewall rules. Remote MCP calls submit inputs to this server; the calculation makes no outbound network requests. Operational events record the tool name, success or error, a controlled error code, and a traffic source classification. Verified automated checks may also record an automation run identifier. Events exclude inputs, results, raw request headers, and automation tokens. Cloudflare may attach platform metadata. Privacy policy: https://packetrove.com/privacy.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • cidr-subtractUse to prepare WireGuard AllowedIPs exceptions or calculate remaining address space relative to supplied include and exclude lists.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Use to prepare WireGuard AllowedIPs exceptions or calculate remaining address space relative to supplied include and exclude lists. Compute union(include) minus union(exclude) as a minimal sorted canonical CIDR list, without adding addresses. Use one address family, a nonempty include list, and at most 1000 entries across both lists, up to 64 characters each. Exclude may be empty. Normalize host bits and count overlaps once. Return cidrs, normalizedInclude, normalizedExclude, and exact decimal-string includedAddressCount, removedAddressCount, remainingAddressCount. Complete removal returns an empty list; more than 10000 output CIDRs returns an error without a partial result. Error issues identify include or exclude and the zero-based entry index. Remote calls submit inputs to this server; the browser calculates locally. This does not inspect live allocation, configure WireGuard, or change firewall rules. Operational events record the tool name, success or error, a controlled error code, and a traffic source classification. Verified automated checks may also record an automation run identifier. Events exclude inputs, results, raw request headers, and automation tokens. Cloudflare may attach platform metadata. Privacy policy: https://packetrove.com/privacy.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • range-to-cidrsUse to prepare an exact CIDR allowlist from one inclusive start/end IPv4 or IPv6 range.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Use to prepare an exact CIDR allowlist from one inclusive start/end IPv4 or IPv6 range. Pass start and end IP addresses of the same family, without CIDR prefixes, at most 64 characters each; end must be at or after start. Return canonical range.first and range.last, minimal sorted cidrs, cidrCount, and exact decimal-string addressCount, without adding addresses. Equal endpoints return one /32 or /128; complete address spaces return /0. Invalid inputs identify the start or end field; reversed endpoints are never swapped. Browser calculations stay local; remote MCP calls submit endpoints to this server. This does not inspect live address usage, modify firewall rules, or export vendor-specific ACLs. Operational events record the tool name, success or error, a controlled error code, and a traffic source classification. Verified automated checks may also record an automation run identifier. Events exclude inputs, results, raw request headers, and automation tokens. Cloudflare may attach platform metadata. Privacy policy: https://packetrove.com/privacy.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
Only reads 2Show 2 read-only tools
  • certificate-bundleUse to inspect a supplied PEM certificate bundle before TLS configuration.No findings

    Why it is in this group

    • The description's first verb is "inspect": a word for reading.
    Description, as the server wrote it
    Use to inspect a supplied PEM certificate bundle before TLS configuration. Accept pem with 1–16 CERTIFICATE blocks and at most 49152 UTF-8 bytes, optional ASCII DNS hostname, and optional zero-based leafIndex. Reject private keys and unsupported blocks without echoing them. Return original certificate positions, Subject, Issuer, Common Name, SANs, validity, CA and Key Usage flags, SHA-256 fingerprints, independently checked candidate links, explicit leaf ambiguity, and stable findings with severity, observed evidence, and nextAction. Verify signatures cryptographically; name matching alone is not verification. Missing selected-leaf issuers and blocked requested hostname checks are warnings; other missing issuers are informational because roots are commonly omitted. All explicitly rejected supplied issuer candidates for the selected leaf produce an error. Unknown checks and key-identifier mismatches alone do not produce that error. A failed candidate does not invalidate another viable link. Verified self-issued CA key rollover is informational; verification with the certificate's own public key remains separate. Hostname checks use DNS SAN with one-label complete leftmost wildcards and no Common Name fallback. Results use this server's evaluation time and do not establish client trust, full RFC 5280 path validity, revocation status, or deployment safety. This remote call transmits certificates and hostname to the server; the website checks locally. Certificate input and details never enter application logs. Do not submit private keys. Operational events record the tool name, success or error, a controlled error code, and a traffic source classification. Verified automated checks may also record an automation run identifier. Events exclude inputs, results, raw request headers, and automation tokens. Cloudflare may attach platform metadata. Privacy policy: https://packetrove.com/privacy.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • public-ipUse to inspect the public IPv4 or IPv6 address observed for the connection making this MCP tool call.Reads text others writeNo findings

    Why it is in this group

    • The description's first verb is "inspect": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Use to inspect the public IPv4 or IPv6 address observed for the connection making this MCP tool call. Takes an empty object and returns ip and family. This is the MCP client connection: a hosted AI client may observe its own exit address, not the user device address. If the user needs their browser or computer connection, direct them to the web tool or a CLI running on that machine. A VPN or proxy changes the observed path. One call observes one address family; it does not discover private local addresses, an address before a proxy, or both address families. The application does not store or log results, and no firewall rules are changed. Cloudflare Worker subrequests can have platform-specific address semantics; the result is not an identity proof. Operational events record the tool name, success or error, a controlled error code, and a traffic source classification. Verified automated checks may also record an automation run identifier. Events exclude inputs, results, raw request headers, and automation tokens. Cloudflare may attach platform metadata. Privacy policy: https://packetrove.com/privacy.

    Annotations the server declares

    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true

This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.

Exfil surface

Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.

No private-data, untrusted-input, outbound path in this server alone: no tool is tagged private-data or outbound.

Reads private data0
Reads files, mail, messages, databases, secrets, calendars or repositories.
Reads text others write2
Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
Can send data out0
Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.

Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.

Remote posture

What anyone on the internet sees at the remote endpoint: an anonymous initialize and tools/list, and the public OAuth metadata.

Tools listed without sign-in

Looked at Nov 20, 2026, 19:04 UTC, unchanged since Oct 22, 2026. No credentials were sent and no tool was called. What is looked at.

Reachable
yes
HTTP status of an anonymous initialize
200
tools/list served without credentials
yes
Tools that change things, listed without credentials
4
Protocol version
2025-06-18
WWW-Authenticate on 401
not observed
resource_metadata in WWW-Authenticate
not observed
RFC 9728 protected resource metadata
not observed
Authorization servers
not observed
Legacy HTTP+SSE transport
no
Dynamic client registration
not observed
Client ID metadata documents
not observed
  • HighThe whole server

    The endpoint served tools/list without credentials, and the list has 4 tools that change things (1 destructive, 3 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://api.packetrove.com/mcp: tools/list returned 6 tools to an anonymous client, including submit-feedback (destructive), cidr-cover (write), cidr-subtract (write), range-to-cidrs (write)

    What this check looks for

Findings

Matches of fixed checks, with the text they matched.

2 info
  • The description mentions "credentials".

    Matched text

    description: "...conversations, raw tool inputs/results, credentials, certificates, logs, or client/session..."

    What this check looks for
  • The description is 2,521 characters long, over the 2,000 this check looks for.

    Matched text

    description ends: "...tools do not depend on feedback. Privacy policy: https://packetrove.com/privacy."

    What this check looks for

History

Changes to its tool definitions and to its remote endpoint, newest first.

No change seen yet. Its tools are rescanned every day, and a remote endpoint is looked at once a week.

Install

Pinned to version 0.3.0. Values in angle brackets are yours to fill in.

claude mcp add --transport http packetrove https://api.packetrove.com/mcp
claude mcp add --transport http packetrove https://api.packetrove.com/mcp

Run in a terminal.

Claude Code permissions

Allow the reads, ask before a change or a way out, deny the deletes. Paste into .claude/settings.json.

.claude/settings.json
{
  "permissions": {
    "allow": [
      "mcp__packetrove__certificate-bundle",
      "mcp__packetrove__public-ip"
    ],
    "ask": [
      "mcp__packetrove__cidr-cover",
      "mcp__packetrove__cidr-subtract",
      "mcp__packetrove__range-to-cidrs"
    ],
    "deny": [
      "mcp__packetrove__submit-feedback"
    ]
  }
}

Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as packetrove. Review it before you use it.

Capability tags behind 2 rules
  • mcp__packetrove__public-ip untrusted-input
  • mcp__packetrove__submit-feedback untrusted-input

Pin

Keep the decision: fail CI when the tools change, follow every change, or show a badge.

Pin in CI

Fail a build when the tools you reviewed are not the tools the server serves today.

curl -s "https://radar.anlyon.com/api/v1/servers/io.github.euyuil/packetrove/verify?hash=76886d32418c0eba5406b76195e0a2e6fe5bf3cfe606240abaea23021eba665d" | jq -e '.changed == false'

Pinned hash 76886d32418c, from Nov 19, 2026, 16:18 UTC.

Watch

Get told when its tools change: a new tool, a new description, a new class.

https://radar.anlyon.com/servers/io.github.euyuil/packetrove/changes.xml

An Atom feed. Works in any feed reader or Slack's RSS app.

Badge

Anlyon Radar badge for io.github.euyuil/packetrove
[![Anlyon Radar](https://radar.anlyon.com/badge/io.github.euyuil/packetrove.svg)](https://radar.anlyon.com/servers/io.github.euyuil/packetrove)

Source: MCP Registry · repository. Synced Oct 21, 2026, 18:33 UTC.

Packetrove MCP server: questions and answers

Answered from the scan above, not from the publisher's description.

Is the Packetrove MCP server safe?

No scan can promise that. What its definitions show: 6 tools, 4 of which can change, send or delete something, and 2 findings from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.

Can the Packetrove MCP server delete data?

Yes. 1 tool can delete, cancel or revoke something: submit-feedback.

Can the Packetrove MCP server do things that cannot be undone?

No tool is classed as irreversible: none of them sends, pays, publishes or merges, going by their definitions.

What tools does the Packetrove MCP server have?

6 tools: 1 destructive, 3 write, 2 read. submit-feedback, cidr-cover, cidr-subtract, range-to-cidrs, certificate-bundle and public-ip.

Does the Packetrove MCP server need sign-in or OAuth?

Its remote endpoint listed its tools to an anonymous client, without sign-in.

Does the Packetrove MCP server send data to outside services?

None of its tools sends, posts or publishes, or takes a recipient, channel or URL.

Has the Packetrove MCP server changed its tools?

Not since Anlyon Radar started tracking it. Last scanned Nov 19, 2026.