backbone is a module that adds in structure to a JavaScript heavy application through key-value pairs and custom events connecting to your RESTful API through JSON There exists a potential Cross Site Scripting vulnerability in the `Model#Escape` function of backbone 0.3.3 and earlier, if a user is able to supply input. This is due to the regex that's replacing things to miss the conversion of things such as `<` to `<`.

engine.io-client is the client for engine.io, the implementation of a transport-based cross-browser/cross-device bi-directional communication layer for Socket.IO. The vulnerability is related to the way that node.js handles the `rejectUnauthorized` setting. If the value is something that evaluates to false, certificate verification will be disabled. This is problematic as engine.io-client 1.6.8 and earlier passes in an object for settings that includes the rejectUnauthorized property, whether it has been set or not. If the value has not been explicitly changed, it will be passed in as `null`, resulting in certificate verification being turned off.

csrf-lite is a cross-site request forgery protection library for framework-less node sites. csrf-lite uses `===`, a fail first string comparison, instead of a time constant string comparison This enables an attacker to guess the secret in no more than (16*18)288 guesses, instead of the 16^18 guesses required were the timing attack not present.

electron-packager is a command line tool that packages Electron source code into `.app` and `.exe` packages. along with Electron. The `--strict-ssl` command line option in electron-packager >= 5.2.1 <= 6.0.0 || >=6.0.0 <= 6.0.2 defaults to false if not explicitly set to true. This could allow an attacker to perform a man in the middle attack.

express-restify-mongoose is a module to easily create a flexible REST interface for mongoose models. express-restify-mongoose 2.4.2 and earlier and 3.0.X through 3.0.1 allows a malicious user to send a request for `GET /User?distinct=password` and get all the passwords for all the users in the database, despite the field being set to private. This can be used for other private data if the malicious user knew what was set as private for specific routes.

console-io is a module that allows users to implement a web console in their application. A malicious user could bypass the authentication and execute any command that the user who is running the console-io application 2.2.13 and earlier is able to run. This means that if console-io was running from root, the attacker would have full access to the system. This vulnerability exists because the console-io application does not configure socket.io to require authentication, which allows a malicious user to connect via a websocket to send commands and receive the response.

marked is an application that is meant to parse and compile markdown. Due to the way that marked 0.3.5 and earlier parses input, specifically HTML entities, it's possible to bypass marked's content injection protection (`sanitize: true`) to inject a `javascript:` URL. This flaw exists because `&#xNNanything;` gets parsed to what it could and leaves the rest behind, resulting in just `anything;` being left.

The airbrake module 0.3.8 and earlier defaults to sending environment variables over HTTP. Environment variables can often times contain secret keys and other sensitive values. A malicious user could be on the same network as a regular user and intercept all the secret keys the user is sending. This goes against common best practice, which is to use HTTPS.

Droppy versions <3.5.0 does not perform any verification for cross-domain websocket requests. An attacker is able to make a specially crafted page that can send requests as the context of the currently logged in user. For example this means the malicious user could add a new admin account under his control and delete others.

restafary is a REpresentful State Transfer API for Creating, Reading, Using, Deleting files on a server from the web. Restafary before 1.6.1 is able to set up a root path, which should only allow it to run inside of that root path it specified.

The riot-compiler version version 2.3.21 has an issue in a regex (Catastrophic Backtracking) thats make it unusable under certain conditions.

A common setup to deploy to gh-pages on every commit via a CI system is to expose a github token to ENV and to use it directly in the auth part of the url. In module versions < 0.9.1 the auth portion of the url is outputted as part of the grunt tasks logging function. If this output is publicly available then the credentials should be considered compromised.

When attempting to allow authentication mode `try` in hapi, hapi-auth-jwt2 version 5.1.1 introduced an issue whereby people could bypass authentication.

i18n-node-angular is a module used to interact between i18n and angular without using additional resources. A REST API endpoint that is used for development in i18n-node-angular before 1.4.0 was not disabled in production environments a malicious user could fill up the server causing a Denial of Service or content injection.

MQTT before 3.4.6 and 4.0.x before 4.0.5 allows specifically crafted MQTT packets to crash the application, making a DoS attack feasible with very little bandwidth.

rails_admin ruby gem <v1.1.1 is vulnerable to cross-site request forgery (CSRF) attacks. Non-GET methods were not validating CSRF tokens and, as a result, an attacker could hypothetically gain access to the application administrative endpoints exposed by the gem.

jshamcrest is vulnerable to regular expression denial of service (ReDoS) when certain types of user input is passed in to the emailAddress validator.

jadedown is vulnerable to regular expression denial of service (ReDoS) when certain types of user input is passed in.

A security issue was found in bittorrent-dht before 5.1.3 that allows someone to send a specific series of messages to a listening peer and get it to reveal internal memory.

A vulnerability was found in the ping functionality of the ws module before 1.0.0 which allowed clients to allocate memory by sending a ping frame. The ping functionality by default responds with a pong frame and the previously given payload of the ping frame. This is exactly what you expect, but internally ws always transforms all data that we need to send to a Buffer instance and that is where the vulnerability existed. ws didn't do any checks for the type of data it was sending. With buffers in node when you allocate it when a number instead of a string it will allocate the amount of bytes.