This package has been deprecated

    Author message:

    No longer maintained


    0.1.5 • Public • Published


    Build Status Coverage Status Code Style Chat Room

    Litespeed is a micro web framework for building APIs in Node.js. Based on configuration and promises, it keeps things fast, simple, predictable, and is a breeze to get started. It comes with built-in input validation, a routing library, an error library, support for pre-handlers (aka middleware), and more. A perfect solution for microservices!

    Why another Node framework?

    There are a lot of great frameworks out there such as Express, Restify, Koa, and Hapi. They have been around for awhile, and each have their place in the Javascript ecosystem. But a lot of the time, they can be overkill for your app, especially if you need something simple but extendable for a microservice API. Litespeed is a lightweight approach that takes ideas from other frameworks and brings them together in a compact and easy to understand way. It's also nice to have built-in validation, error handling, and logging ;)


    $ npm install litespeed --save
    # or to use the CLI 
    $ npm install litespeed -g


    const Litespeed = require('litespeed')
    new Litespeed({/* configuration */}).route({
      method: 'GET',
      url: '/hello/:name',
      handler: (req) => {
        return { message: `Hello, ${}!` }

    That's it! Run this code and go to http://localhost:8000/hello/jason. You should be greeted with {"message":"Hello, jason!"}.


    There are several config options, all of which have optimal values by default. But they can be easily changed by passing an object to the Litespeed instance (see the example above).

    • name The name of the server, set in the Server response header. Can be set to false to omit the header entirely.

      Type: string, boolean
      Default: Litespeed

    • host The host to run the server on. Uses the HOST environment variable if it exists.

      Type: string
      Default: process.env.HOST || ''

    • port The port to run the server on. Uses the PORT environment variable if it exists.

      Type: number
      Default: process.env.PORT || 8000

    • trailingSlash Whether to accept a URL both with and without a trailing slash. If true, the trailing slash is ignored and the route is still found. If false, the route will return a 404 with a trailing slash.

      Type: boolean
      Default: true

    • pretty Whether to prettify the response JSON.

      Type: boolean
      Default: false or true if NODE_ENV starts with "dev"

    • timeout The amount of time in milliseconds to wait before a request is timed out, returning a 408 error to the client.

      Type: number
      Default: 5000 (5s)

    • payloadLimit The max size (in bytes) a request payload can be. This helps prevents malicious use by client's sending giant payloads to your API. Will return a 413 error if exceeded.

      Type: number
      Default: 1048576 (1mb)

    • stripUnknown Whether to take out unknown values from the request payload and URL query data. This only takes effect if validation is running on the endpoint, and will throw away any value not included in the validation object.

      Type: boolean
      Default: true

    • protective Whether to add basic security headers on the response and require a user-agent. Will currently add X-Content-Type-Options: nosniff and X-Frame-Options: deny, and send a 403 error if no User-Agent header is present on the request.

      Type: boolean
      Default: true

    • realIp Whether to check for an IP address in the X-Forwarded-For header, taking the first address from the comma-separated list. This helps get the client's real IP address if the request is coming through a proxy server. Will fallback to the actual IP address if the header is not found.

      Type: boolean
      Default: true

    • logs Tag names of the loggers to enable. See Logging for more info on the tags available. Can also be set to false to disable all logging.

      Type: array, boolean
      Default: all

    • logTimestamp Whether to log the current timestamp with each message.

      Type: boolean
      Default: true

    • logColors Whether to use CLI logColors when logging.

      Type: boolean
      Default: true

    • preHandlers Global functions to run for every route in the app. See PreHandlers for more info.

      Type: array
      Default: []

    Routes Config

    A route is defined by a simple object with the following configuration.

    • method The method to use for the route. This is usually one of HEAD, GET, POST, PUT, PATCH, DELETE if using the HTTP protocol, but can be set to anything if you're using a custom protocol.

      Type: string

    • url The endpoint of the route. This can include segments or be a regular expression. See Router for details.

      Type: string, regexp

    • statusCode The HTTP status code to return on a successful response.

      Type: number
      Default: 200

    • validate The validations to run against the request. See Validation for more info.

      Type: object

    • preHandlers The preHandler functions to run before the route's handler. See PreHandlers for more info.

      Type: array

    • handler The request handler function. See Handlers for more info.

      Type: function

    • onError A custom function to run when an error occurs (overwrites default error handling). See Errors for more info. Note: if this is being used, you are responsible for your own error logging!

      Type: function


    The HTTP OPTIONS request is handled automatically for each URL. If requested, the server will respond with an empty body and the Allow header with each method supported for that particular URL.


    Route handlers are functions on the route config that handle sending responses to the client. To respond from the handler, you can either return a value or return a Promise. The function has two parameters passed to it: the request object and the response object.

      // ...
      handler: (request, response) => {
        return 'welcome!'


    The request parameter is an object that contains information about the client request.

    • body The payload data from the request body.
    • query The query data from the request URL.
    • params The param data from the request URL. See Router for more info.
    • headers The request headers.
    • context The context data passed through the preHandlers. See Plugins for more info.
    • info Metadata about the request.
      • remoteAddress The client's IP address


    Unlike other Node web frameworks, the response parameter is not used to actually send the response itself--that is done by returning from the handler (with the exception of .redirect). Instead, this object contains helper methods to mutate the response before it is sent.

    • .pass(name, value) Passes context between preHandlers (see PreHandlers for more info). Context data is accessed in the handler with request.context.*.
    • .setHeader(name, value) Sets a response header.
    • .removeHeader(name, value) Removes a response header.
    • .redirect(url, code) Sends a redirect response to the client. code defaults to 301 if not set. There is no need to return anything from the handler after this is called.


    Since handlers are based on Promises, you can easily use ES7's async/await feature. Any errors thrown within the handler are caught by Litespeed's error handler and outputted correctly. This removes the need to have a bunch of try/catch blocks (though you can still have them if you need to). Errors within sub-promises (such as the createUser function in the example below) will bubble up to the handler and outputted the same way.

      method: 'POST',
      url: '/signup',
      statusCode: 201,
      handler: async (req) => {
        const user = await createUser(req.body)
        return { user }

    Note: If using Node >=6, all you need for async/await support is transform-async-to-generator rather than an entire preset.


    Route preHandlers follow the exact same structure as Handlers, but you can pass an array of them to the routes config as well as the server config. Note: they are guaranteed to run in order, making it possible to rely on context from other preHandlers!


    Litespeed comes with a built-in router that supports segment parameters and regular expressions. This provides a quick and flexible way to define your endpoints. Note: You cannot have two routes defined with the same method and URL or a startup error will be thrown.


    Segments are named sections of the URL prefixed with :, and they can pass values to the handler. You can have as many segments as you want in your URL.

      // ...
      url: '/welcome/:name',
      handler: (request) => {
        // if requested with /welcome/jason
        // expect to equal 'jason'

    Regular Expressions

    The router also supports regular expressions, including their capture groups for getting data. If you define a URL as a regex with a capture group (a piece wrapped in parentheses), you can access that data by its regex index ($1, $2, etc.) Note: if using a regex as a URL, make sure you either use a regular expression literal or new RegExp(). Read this if you're new to regular expressions in Javascript.

      // ...
      url: /^\/welcome\/(.*)$/,
      handler: (request) => {
        // if requested with /welcome/jason
        // expect request.params.$1 to equal 'jason'


    A validate object can be used on the routes config to run validations against the request body, query, and params. Litespeed comes with a list of predefined validations, which can be chained together. See the example below on how to use. If stripUnknown is enabled on the server, any values not in the validate object will be throw away.

    The response is not sent after one failed validation. Litespeed will continue through the validations and display all errors in the response rather than just the first one. However, only one error from each validation chain is thrown at a time.

    const { Validator } = require('litespeed')
      url: '/resource/:id'
      validate: {
        params: {
          id: new Validator().isUUID()
        body: {
          name: new Validator().required(),
          email: new Validator().required().isEmail()

    Validation errors will look like this in the response:

      "statusCode": 400,
      "error": "Bad Request",
      "message": "A validation error occurred",
      "validation": [
        { "field": "id", "message": "...", "where": "params" },
        { "field": "name", "message": "...", "where": "body" }

    Predefined Validations

    • .required()
    • .contains(string)
    • .equals(compare)
    • .isAfter(date)
    • .isAlpha(locale)
    • .isAlphanumeric(locale)
    • .isAscii()
    • .isBase64()
    • .isBefore(date)
    • .isBoolean()
    • .isCurrency(opts)
    • .isDataURI()
    • .isDate()
    • .isDivisibleBy(num)
    • .isEmail(opts)
    • .isFQDN(opts)
    • .isIpAddress(ver)
    • .isEnum(values)
    • .isNumber(opts = {})
    • .isJSON()
    • .isLength(opts)
    • .isLowercase()
    • .isMACAddress()
    • .isPhoneNumber(locale)
    • .isMongoId()
    • .isURL(opts)
    • .isUUID(ver)
    • .isUppercase()
    • .isWhitelisted(chars)
    • .matches(pattern, mods)

    Most of these validations use validator.js, so visit the Readme in that repo for further details.


    Any errors occurring throughout the application or in an async/await function will be automatically logged and formatted for the response. Errors can occur in a few different ways:

    • By throwing anywhere in the app
    • Runtime code errors
    • Returning an error from a handler
    • An unhandled promise rejection

    Error responses are simply objects with the following structure:

    • statusCode The status code of the error (uses 500 for unknown errors)
    • error The error type that occurred
    • message A descriptive error message for the client

    If a Javascript Error is thrown, a 500 Internal Server Error will be returned, and if in dev mode (meaning NODE_ENV starts with dev), the error message will be sent to the response as well for easy debugging. All server errors (meaning statusCode is >= 500) will be logged to the console with a description and a stack trace.

    Litespeed comes with a list of predefined errors that accept custom messages (set as the message key in the response). See the example below on how to use them.

    const { Errors } = require('litespeed')
      handler: async () => {
        const resource = getNonExistantResource()
        if (!resource) {
          throw new Errors().notFound('resource does not exist')
        return resource
      "statusCode": 404,
      "error":"Not Found",
      "message":"resource does not exist"

    Predefined Errors

    • 400 .badRequest()
    • 401 .unauthorized()
    • 402 .paymentRequired()
    • 403 .forbidden()
    • 404 .notFound()
    • 405 .methodNotAllowed()
    • 406 .notAcceptable()
    • 407 .proxyAuthRequired()
    • 408 .requestTimeout()
    • 409 .conflict()
    • 410 .gone()
    • 411 .lengthRequired()
    • 412 .preconditionFailed()
    • 413 .payloadTooLarge()
    • 414 .uriTooLong()
    • 415 .unsupportedMediaType()
    • 416 .rangeNotSatisfiable()
    • 417 .expectationFailed()
    • 422 .unprocessableEntity()
    • 423 .locked()
    • 424 .failedDependency()
    • 426 .upgradeRequired()
    • 428 .preconditionRequired()
    • 429 .tooManyRequests()
    • 431 .headersTooLarge()
    • 500 .internal()
    • 501 .notImplemented()
    • 502 .badGateway()
    • 503 .serviceUnavailable()
    • 504 .gatewayTimeout()
    • 505 .httpVersionNotSupported()
    • 507 .insufficientStorage()
    • 508 .loopDetected()
    • 509 .bandwidthLimitExceeded()
    • 510 .notExtended()
    • 511 .networkAuthRequired()


    By default, all Litespeed logging is turned on. This can be modified by specifying log tags in the server config. Each line outputted to the console includes the server name, a timestamp, and the message. Here are the available log tags:

    • server Outputs a message when the server starts with its URL.
    • request Outputs the method, URL, IP address, and response code when a client makes a request.
    • error Outputs any server errors, meaning any error with a statusCode >= 500 (which includes runtime errors). Logged errors will include a message, and a stack trace.

    Colors in the output can be turned off if they are causing problems by specifying logColors: false in the server config. Note: if you specify any log tags, the default tags will be overwritten. So whatever you specify will be the only active tags.

    Server API

    The Litespeed server is a class that must be instantiated with the new keyword, and can be passed an optional object of configuration options.

    const server = new Litespeed({/* configuration */})

    Starts the server on the host and port defined in the instantiation. This will return a promise or you can use a callback, both of which pass the URL of the server.

    /* with a callback */
    server.start((url) => console.log(`Running at ${url}`))
    /* or with a promise */
    server.start().then((url) => console.log(`Running at ${url}`))

    Creates a new route. Can be a single route config, or an array of configs for creating multiple routes at once. See Routes Config for structure.

    /* for a single route */
    server.route({/* route config */})
    /* for many routes */
      {/* route config */},
      {/* route config */}

    You can use this to scan a directory and pull in routes from many files. Prevents having to manually require and create all your routes. Takes a config object with the following structure.

    • dir A glob pattern to search for.

      Type: string
      Default: routes/**/*.js

    • cwd The current working directory to use.

      Type: string
      Default: process.cwd()

      dir: 'routes/**/*.js',
      cwd: __dirname

    Lets you inject API requests without the overhead of starting the server. This works great for running integration tests on your endpoints. Takes an object with a method and url, as well as optional headers and body. Returns a Promise resolving with the response.

    /* define the route */
      method: 'POST',
      url: '/test',
      handler: (req) =>
    /* test the route */
      method: 'POST',
      url: '/test',
      body: { name: 'Jason' }
    }).then((res) => {
      // expect res to equal 'Jason'


    Litespeed easily supports third-party plugins through the use of PreHandlers. To create a plugin, simply create a preHandler that exports a function returning a Promise (if async). Then it can be set as a global preHandler in the app using it.

    You can also pass context from your plugin to other plugins or to the route handler by using response.pass(name, value) (see response). These values are then accessed with request.context.*.

    Popular Plugins


    If you come across an issue or have a feature idea, don't hesitate to create a pull request/issue to discuss it.

    Just do it.




    npm i litespeed

    DownloadsWeekly Downloads






    Last publish


    • jsonmaur