Use alongside eslint-plugin-security
Keep eslint-plugin-security and add Interlace next to it. Measured on 100 vulnerable and 96 safe cases, what each one finds that the other doesn't.
If your project already runs eslint-plugin-security, keep it. This page shows what adding Interlace next to it finds, measured, and the config that does it.
The config
Two imports and two lines in the eslint.config.js you already have:
import security from 'eslint-plugin-security';
import secureCoding from 'eslint-plugin-secure-coding';
import nodeSecurity from 'eslint-plugin-node-security';
export default [
security.configs.recommended,
secureCoding.configs.recommended,
nodeSecurity.configs.recommended,
];npm install --save-dev eslint-plugin-secure-coding eslint-plugin-node-securityThose two cover general Node.js code. If you use Express, a browser front end, PostgreSQL, JWTs or MongoDB, add the plugin for it too. The table below shows which plugin found what.
What the measurement shows
Both were run over the same corpus: 100 vulnerable and 96 safe code samples across 45 CWEs. Interlace ran with the recommended preset of each of its security plugins.
| Vulnerable samples (100) | Count |
|---|---|
| Found by Interlace only | 74 |
| Found by both | 14 |
Found by eslint-plugin-security only | 1 |
| Found by neither | 11 |
| Safe samples (96) flagged anyway | Count |
|---|---|
| Interlace | 0 |
eslint-plugin-security | 12 |
Which Interlace plugin found the 74
A sample can be found by more than one plugin, so these add up to more than 74.
| Plugin | Samples | Examples |
|---|---|---|
node-security | 19 | weak Diffie-Hellman and ECDH keys, TLS checks turned off, SSRF |
secure-coding | 18 | hardcoded credentials, log injection, failing open, homoglyphs |
browser-security | 13 | incomplete URL sanitizing, postMessage to * |
express-security | 13 | stack traces in responses, host-header poisoning in reset links |
mcp-sdk-security | 9 | MCP token passthrough, DNS rebinding |
postgresql-security | 4 | SQL built from strings |
jwt-security | 4 | none algorithm, tokens with no issued-at |
mongodb-security | 2 | NoSQL operator injection |
The one only eslint-plugin-security found
security/detect-disable-mustache-escape flags Mustache templates with escaping turned off. No Interlace rule catches that sample today, and that is one reason to keep eslint-plugin-security installed.
When both report the same line
14 samples are found by both, so on those lines you will see two messages. If you prefer one, turn off the matching security/* rule after you have read both messages.
What this does and doesn't show
- We wrote this corpus. It measures the cases our rules were built for, so it says more about what Interlace covers than about how often each tool is right on your code.
- On real code, both are noisy. On 20 open-source projects (3.04M lines), Interlace's
recommendedfindings were right 28.6% of the time andeslint-plugin-security's 13.0% (24 findings sampled per side). See benchmarks. - Reproduce it:
npm run ilb:cwe-corpusin ofri-peretz/eslint. The run behind this page isbenchmarks/results/ilb-cwe-corpus/2026-11-15.json, againsteslint-plugin-security4.2.1.
Building secure JavaScript with Interlace? Star the repo to get new rules and CWE coverage as we ship them — or follow the AI-code-security benchmarks behind them.