More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    railguard: Honest Reviewer for Node.js & Express

    1

    Makes your AI agent stop secret leaks, test edge cases before calling code done, and give an honest yes/no before you ship a Node.js or Express app. Powered by the free, zero-dependency railguard npm package.

    Free

    Security scanned
    railguard: Honest Reviewer for Node.js & Express

    railguard: Honest Reviewer for Node.js & Express

    Example session with this skill installed

    I'm about to commit my Express API changes. Can you check my .env setup and the new slugify utility?

    • Read your context and instructions
    • Compiled the railguard

    I have reviewed your environment configuration and the slugify utility. Your .env file is currently ignored by git, but your env.schema.js is missing the new CLOUDINARY_API_SECRET you added.

    Security and Edge Case Audit

    CategoryStatusDetails
    Secrets ScanPASSnpx railguard scan returned 0 findings.
    Git HygienePASS.env is correctly listed in .gitignore.
    Env SchemaFAILCLOUDINARY_API_SECRET is missing from env.schema.js.
    Edge CasesFAILprobe found a crash on null input in slugify.js.
    Route ValidationWARNPOST /api/media uses req.body without validate() middleware.

    Recommended Fixes

    // 1. Update env.schema.js
    const { t } = require('railguard');
    module.exports = {
      // ... existing keys
      CLOUDINARY_API_SECRET: t.string().secret().describe('Cloudinary private key')
    };
    
    // 2. Fix slugify.js crash
    // The probe failed with: ✗ crash (TypeError: Cannot read properties of null)
    export const slugify = (str) => {
      if (typeof str !== 'string') return ''; // Add guard clause
      return str.toLowerCase().trim().replace(/[^\w\s-]/g, '').replace(/[\s_-]+/g, '-');
    };
    

    Next steps

    1. Run npx railguard init to sync your schema with the new environment variables.
    2. Apply validate({ body: { ... } }) to your new Express route to block mass-assignment.
    3. Run npx railguard hook to ensure these checks happen automatically on every commit.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    AI agents write code fast and tell you it's done. This skill makes your agent act like an honest senior reviewer on Node.js and Express projects:

    Stops secret leaks. Paste an API key into the chat and the agent tells you to rotate it, without repeating it back. It keeps secrets out of code and runs a secret scan before every commit. Checks edge cases. Before calling a function done, it runs railguard probe, which throws dozens of tricky inputs (null, empty strings, NaN, invalid dates, emoji, proto payloads) at your code and reports exactly what crashed. Validates your .env at startup. Missing or mistyped config stops the app at boot with a clear message. Values are never printed. Builds safer Express APIs. Validated input, one JSON error format, no stack traces sent to users, and ownership checks. Gives an honest pre-ship check. Ask "is it ready?" and it walks a 5-point checklist, marking each item as passed, failed or not checked.

    The skill uses railguard, a free, MIT-licensed npm package with zero dependencies. It works with Claude Code, Cursor, Codex and any agent that reads SKILL.md.

    Tested: in 3 scenarios on a sample Express app with problems planted in it (a pasted API key, buggy date helpers, "is it ready to ship?"), the agent rotated and never repeated the key, refused to call broken code done, and gave a clear "no" on an unsafe deploy. It also found a timezone bug the automated tests couldn't see.

    Tags

    nodejs, express, security, secrets, env, dotenv, testing, edge-cases, code-review, mern, javascript, pre-commit

    Requires Node v20.19+. Only probe utility modules without side effects. Does not replace human security reviews for auth or payment logic.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 7 days ago

    • Free with an Agensi account

    Requires

    Node.js
    Git

    Permissions

    Terminal / Shell
    Write Files

    File Scopes

    .env*
    .gitignore
    env.schema.js
    .git/hooks/pre-commit
    AGENTS.md
    CLAUDE.md

    Reads .env only to validate keys. Values are never printed, and scan output is masked. Writes files only when you ask: env.schema.js (init), a pre-commit hook (hook), and a rules block in AGENTS.md/CLAUDE.md (ai). No network access, except npm install if you approve it.

    Listed7 days ago

    What's inside

    Frequently Asked Questions

    Railguard acts as a proactive security layer that detects exposed credentials, validates environment variables against a schema, and stress-tests your JavaScript functions for edge-case crashes before they reach production.

    Railguard is designed for Node.js environments (v20.19+) and integrates seamlessly with any AI agent capable of executing CLI commands or modifying JavaScript files, specifically supporting Express 4 and 5.

    The 'probe' feature automatically executes your utility functions with malformed inputs, null values, and boundary cases to identify potential prototype pollution or unhandled exceptions that manual testing often misses.

    Setup is automated: simply run `npx railguard init` to generate your environment schema and `npx railguard hook` to install the pre-commit secret scanners.

    You receive the full Railguard skill package, including the CLI toolkit, Express middleware components, and a suite of pre-configured audit checklists for production readiness.

    The 'probe' command executes the code it tests, so it should only be used on utility modules without side effects; it is not intended for testing live database connections or UI components.

    New in LLM Ops