Generate IDs, QR codes, and hashes, encode values, geolocate IPs, plus gated host diagnostics.
Generate random IDs, QR codes, and hashes, encode and decode values, and geolocate IPs, plus gated network and system diagnostics, via MCP. STDIO or Streamable HTTP.
Public Hosted Server: https://toolkit.caseyjhand.com/mcp
A standalone developer-utilities server — the five always-on tools need no upstream API: generate identifiers, QR codes, and cryptographic digests, encode and decode values, and geolocate a public IP or hostname. Two more tools report diagnostics about the server's own host, gated off by default. Runs as a stdio process, a local Streamable HTTP server, or the public hosted endpoint above.
| Tool | Description |
|---|---|
toolkit_hash_value | Generate a cryptographic digest (sha256/sha512/sha1/md5), or constant-time-compare a value against an expected digest. |
toolkit_generate_id | Mint cryptographically-random identifiers — UUIDv4, UUIDv7, or ULID — singly or in batches up to 1000. |
toolkit_generate_qr | Encode text or a URL into a QR code as SVG markup, base64 PNG, or a terminal-renderable string. |
toolkit_encode_value | Encode or decode a value across base64, base64url, hex, or URL percent-encoding, in either direction. |
toolkit_geolocate_ip | Resolve a public IP or hostname to geographic and network metadata — country, city, coordinates, ASN, timezone. |
toolkit_check_network | Gated, off by default. Read-only network diagnostics from the server host — ping, traceroute, TCP connectivity, or egress-IP detection. |
toolkit_check_system | Gated, off by default. Report a facet of the server host's system state — OS, CPU, memory, load average, or network interfaces. |
toolkit_hash_value tooloperation: generate (lowercase-hex digest) or compare (timing-safe check via timingSafeEqual)sha256 (default) and sha512 for security; sha1 and md5 are exposed for checksum and file-integrity compatibility only — never for passwords or signaturesinputEncoding reads value as utf8 (default), hex, or base64, so binary blobs skip a decode round-triptoolkit_generate_id tooltype: uuid_v4 (random, default), uuid_v7 (time-ordered, sortable by creation), or ulid (26-char Crockford base32, lexicographically sortable)count mints a batch up to 1000 in one call; the returned ids array always holds exactly count valuesuuid_v7 and ulid batches are monotonic — strictly increasing even within the same millisecond — so ids stays in sorted creation ordertoolkit_generate_qr toolformat: svg (inline markup), png_base64 (raster bytes with mimeType and byteLength), or terminal (Unicode block string)errorCorrection (L/M/Q/H) trades data capacity for damage tolerance; margin sets the quiet-zone width; scale sets pixels per module for raster outputversion (1–40) reflects how dense the encoded data ispng_base64 also arrives as an MCP image content block, so a client reading content[] can render the code without decoding structuredContent(modules + 2 × margin) × scale — so a dense symbol at a high scale is rejected with a typed raster_too_large error naming a scale that fits; svg and terminal are unboundeddata is capped at 2953 bytes — the absolute ceiling (version 40, level L, byte mode); usable capacity is lower at higher errorCorrection levels, so over-capacity input is rejected with a typed data_too_large error rather than a generic failuretoolkit_encode_value toolencoding: base64, base64url (URL-safe alphabet), hex, or url (percent-encoding)operation: encode (raw UTF-8 → encoding) or decode (encoded value → text)decode_failed error with a recovery hint, not a silent best-efforttoolkit_geolocate_ip toolproxy, hosting, and mobile flag when the address is a proxy/VPN/Tor exit, a datacenter network, or a mobile carrier — a true on any of them means the coordinates describe infrastructure, not a person. Absent when the provider doesn't report themresolvedIp echoes the IP actually located, and source names the answering providerorg, isp, as) cannot flood or format a model's contextTOOLKIT_GEO_BASE_URL); results are cached in memory by resolved IP under a fixed entry captoolkit_check_network toolTOOLKIT_ENABLE_NET_DIAGNOSTICS=true; absent from tools/list otherwisemode: ping (ICMP round-trip), traceroute (hop path to the target), connectivity (raw TCP connect to target on port), or public_ip (the host's own egress IP)reachable: false — a valid result, not an errorTOOLKIT_ALLOW_PRIVATE_NETWORK=true, which keeps the cloud-metadata endpoint blocked by defaulttoolkit_check_system toolTOOLKIT_ENABLE_SYSTEM_INFO=true; absent from tools/list otherwisewhat: os, cpu, memory, load, or interfaceswhatos and interfaces disclose host topology and version detailsBuilt on @cyanheads/mcp-ts-core: stdio and Streamable HTTP transports, pluggable auth (none / jwt / oauth), swappable storage (in-memory, filesystem, Supabase, Cloudflare KV/R2/D1), structured logging with optional OpenTelemetry tracing.
Toolkit-specific:
node:crypto and the qrcode library; no upstream callstoolkit_geolocate_ip is the one keyless-by-default network call (ip-api free tier, optional TOOLKIT_GEO_API_KEY); the server calls the provider directly and re-checks the DNS-resolved IP against private ranges, so a hostname can't smuggle a request to an internal addresstoolkit_check_network, toolkit_check_system) are absent from tools/list unless explicitly enabled, so a hosted instance exposes no SSRF or info-disclosure surface by defaultdata capped at 2953 bytes, rendered PNGs capped at 2048 px per side, ID batches capped at 1000; CSPRNG-backed primitives with constant-time hash comparison via timingSafeEqualAgent-friendly output:
resolvedIp (the IP actually located) and source (the answering provider); absent upstream fields are reported as unknown, never inventedorg, isp, as) are length-bounded and stripped of control characters before they reach the response, so untrusted registry text can't flood or format a model's contextoperation, format, mode, and what fields echo back exactly what ran, with only the branch-relevant fields populated per call; an unreachable host in toolkit_check_network reports reachable: false as valid data, not an errorreason plus a next-step recovery hint (e.g. decode_failed, raster_too_large, private_target_blocked)A public instance is available at https://toolkit.caseyjhand.com/mcp — no installation required. Point any MCP client at it via Streamable HTTP:
{
"mcpServers": {
"toolkit-mcp-server": {
"type": "streamable-http",
"url": "https://toolkit.caseyjhand.com/mcp"
}
}
}
Add the following to your MCP client configuration file. No API key is required — the five always-on tools and the default keyless geolocation tier work out of the box.
{
"mcpServers": {
"toolkit-mcp-server": {
"type": "stdio",
"command": "bunx",
"args": ["@cyanheads/toolkit-mcp-server@latest"],
"env": {
"MCP_TRANSPORT_TYPE": "stdio",
"MCP_LOG_LEVEL": "info"
}
}
}
}
Or with npx (no Bun required):
{
"mcpServers": {
"toolkit-mcp-server": {
"type": "stdio",
"command": "npx",
"args": ["-y", "@cyanheads/toolkit-mcp-server@latest"],
"env": {
"MCP_TRANSPORT_TYPE": "stdio",
"MCP_LOG_LEVEL": "info"
}
}
}
}
Or with Docker:
{
"mcpServers": {
"toolkit-mcp-server": {
"type": "stdio",
"command": "docker",
"args": ["run", "-i", "--rm", "-e", "MCP_TRANSPORT_TYPE=stdio", "ghcr.io/cyanheads/toolkit-mcp-server:latest"]
}
}
}
To enable the gated host-probing tools, add their flags to env (or -e for Docker):
"env": {
"MCP_TRANSPORT_TYPE": "stdio",
"TOOLKIT_ENABLE_NET_DIAGNOSTICS": "true",
"TOOLKIT_ENABLE_SYSTEM_INFO": "true"
}
For Streamable HTTP, set the transport and start the server:
MCP_TRANSPORT_TYPE=http MCP_HTTP_PORT=3010 bun run start:http
# Server listens at http://localhost:3010/mcp
git clone https://github.com/cyanheads/toolkit-mcp-server.git
cd toolkit-mcp-server
bun install
Every variable is optional. Server-specific options are validated at startup via the Zod schema in src/config/server-config.ts.
| Variable | Description | Default |
|---|---|---|
TOOLKIT_ENABLE_NET_DIAGNOSTICS | Register the gated toolkit_check_network tool. Leave off for hosted or shared deployments. | false |
TOOLKIT_ENABLE_SYSTEM_INFO | Register the gated toolkit_check_system tool. Meaningful only on a local or self-hosted deployment. | false |
TOOLKIT_ALLOW_PRIVATE_NETWORK | With network diagnostics on, permit private/reserved/loopback targets. The second explicit gate. | false |
TOOLKIT_GEO_API_KEY | API key for the geolocation endpoint, if it requires one. | none |
TOOLKIT_GEO_BASE_URL | Base URL for an ip-api-compatible geolocation endpoint. The default is plaintext HTTP — ip-api's HTTPS endpoint is not part of the keyless free tier and answers 403 SSL unavailable for this endpoint without a paid key. Point this at an HTTPS endpoint (with TOOLKIT_GEO_API_KEY) to encrypt the provider request. | http://ip-api.com |
TOOLKIT_GEO_CACHE_TTL_SECONDS | In-memory geolocation cache TTL in seconds. | 3600 |
TOOLKIT_GEO_RATE_LIMIT_PER_MIN | Max geolocation requests per minute. | 45 |
MCP_TRANSPORT_TYPE | Transport: stdio or http. | stdio |
MCP_HTTP_PORT | Port for the HTTP server. | 3010 |
MCP_SESSION_MODE | auto, stateful, or stateless. No tool requests multi-round input. auto is the framework schema default and resolves to stateful, but with MCP_SESSION_MODE unset the server resolves stateless from createApp({ sessionMode }); an explicit MCP_SESSION_MODE value still overrides it. | stateless |
MCP_AUTH_MODE | Auth mode: none, jwt, or oauth. | none |
MCP_LOG_LEVEL | Log level (RFC 5424). | info |
OTEL_ENABLED | Enable OpenTelemetry instrumentation (spans, metrics, completion logs). | false |
See .env.example for the full list of optional overrides.
Build and run:
# One-time build
bun run rebuild
# Run the built server
bun run start:stdio
# or
bun run start:http
Run checks and tests:
bun run devcheck # Lint, format, typecheck, security, changelog sync
bun run test # Vitest test suite
bun run lint:mcp # Validate MCP definitions against spec
docker build -t toolkit-mcp-server .
docker run --rm -e MCP_TRANSPORT_TYPE=http -p 3010:3010 toolkit-mcp-server
The Dockerfile defaults to HTTP transport, stateless session mode, and logs to /var/log/toolkit-mcp-server. OpenTelemetry peer dependencies are installed by default — build with --build-arg OTEL_ENABLED=false to omit them.
| Directory | Purpose |
|---|---|
src/index.ts | createApp() entry point — registers tools and inits services, with fail-closed gating for the two host-probing tools. |
src/config | Server-specific environment variable parsing and validation with Zod. |
src/mcp-server/tools | Tool definitions (*.tool.ts). Seven tools — five always-on, two gated. |
src/services/geo | Geolocation service — DNS resolution, provider call with retry/backoff, normalization, in-memory cache. |
src/services/network | Network-diagnostic service plus the shared target validator and private-range classifier. |
tests/ | Unit and integration tests mirroring the src/ structure. |
See CLAUDE.md / AGENTS.md for development guidelines and architectural rules. The short version:
try/catch in tool logicctx.log for request-scoped logging, ctx.state for tenant-scoped storagecreateApp() arrays in src/index.tsIssues are welcome. Run checks and tests before submitting:
bun run devcheck
bun run test
Apache-2.0 — see LICENSE for details.
Source-derived launch command. Check the maintainer’s required arguments and credentials before running:
npx -y @cyanheads/toolkit-mcp-serverMerge this template into ~/Library/Application Support/Claude/claude_desktop_config.json. Keep existing servers. Add any arguments, credentials, and permissions required by the maintainer; this template has not been install-tested.
{
"mcpServers": {
"io-github-cyanheads-toolkit-mcp-server": {
"command": "npx",
"args": [
"-y",
"@cyanheads/toolkit-mcp-server"
]
}
}
}Restart Claude Desktop completely for changes to take effect. Confirm the server appears connected in the client’s tool list, then try a read-only example from its documentation.
Claude Desktop setup referenceio.github.cyanheads/toolkit-mcp-server works with any MCP-compatible client. Copy the config snippet from the Configuration section above and add it to the file shown for your client, then restart the application.
~/Library/Application Support/Claude/claude_desktop_config.jsonRestart Claude Desktop completely for changes to take effect.~/.cursor/mcp.jsonRestart Cursor for changes to take effect..vscode/mcp.jsonReload VS Code window for changes to take effect.~/.codeium/windsurf/mcp_config.jsonRestart Windsurf for changes to take effect..mcp.jsonSave at the project root, then start Claude Code in that project and review the MCP server approval prompt. Keep real credentials out of shared files.