Korext: AI Code Governance

korext/enforce-action

Connects Claude to Korext's policy enforcement engine with 72 compliance packs and 532 rules across 13 languages. Exposes operations to scan code against policy sets like PCI DSS, OWASP, GDPR, and HIPAA, then retrieve violations as structured results with proof bundles. Built for AI coding workflows where you need to validate generated code against security and compliance standards before it ships. The MCP interface wraps their enforcement API that normally runs in GitHub Actions, so you can check policies interactively during development instead of waiting for CI. Supports regional data processing and cryptographic signing of audit trails.

tools: 5
auth: Secret needed
transport: STDIO, HTTP
last updated: Apr 15, 2026

View on GitHub

Install

From registry metadata

$ npx add-mcp https://app.korext.com/api/mcp --name governance --header 'Authorization: Bearer ${API_TOKEN}'

Run in your terminal. add-mcp asks which agents to add it to, then asks for each value it needs. Add -g to install it for every project.

Review the command, arguments and environment values before installing. MCP servers run with your local permissions.

Other ways

Claude Code

$ claude mcp add --transport http governance https://app.korext.com/api/mcp --header 'Authorization: YOUR_AUTHORIZATION'

Run in your terminal. Add --scope user to make it available in every project.

Tools

Verified live · 5 tools

Verified live against the running server on Nov 6, 2026.

get_directivesGet enriched policy directives and living standards for one or more policy packs. Returns rules with governance context (CWE, MITRE, regulatory references) for inference time enforcement.1 params

Get enriched policy directives and living standards for one or more policy packs. Returns rules with governance context (CWE, MITRE, regulatory references) for inference time enforcement.

Parameters* required

packIdvalue

Policy pack ID or array of IDs (e.g. 'web', ['web', 'pci-dss-v1']). Defaults to 'web'.

check_codeCheck a code snippet against one or more policy packs. Returns violations with severity, governance context (CWE, MITRE, regulatory), and line numbers.2 params

Check a code snippet against one or more policy packs. Returns violations with severity, governance context (CWE, MITRE, regulatory), and line numbers.

Parameters* required

code*string

The code snippet to check.

packIdvalue

Policy pack ID or array of IDs. Defaults to 'web'.

list_packsList available policy packs with industry and region tags.2 params

List available policy packs with industry and region tags.

Parameters* required

industrystring

Filter by industry (e.g. finance, healthcare, defense, aerospace, energy, technology)

regionstring

Filter by region (e.g. us, eu, uk, global)

generate_proofScan code against policy packs and generate a cryptographically signed proof bundle. Returns the bundle ID, decision (PASS/BLOCK), violation count, and verification URL.4 params

Scan code against policy packs and generate a cryptographically signed proof bundle. Returns the bundle ID, decision (PASS/BLOCK), violation count, and verification URL.

Parameters* required

code*string

The code to scan.

fileNamestring

File name for context

languagestring

Programming language (typescript, python, etc.)

packIdvalue

Policy pack ID(s). Defaults to 'web'.

get_proofRetrieve an existing proof bundle by ID. Returns the decision, violations, packs, and verification status.1 params

Retrieve an existing proof bundle by ID. Returns the decision, violations, packs, and verification status.

Parameters* required

bundleId*string

The proof bundle ID (kpb_...)

Readme

KOREXT Enforce Action

Enforce compliance policies on AI generated code in your GitHub workflows.

72 policy packs. 532 rules. 13 languages. Violations appear as GitHub Code Scanning annotations on pull requests.

Quick Start

Add this to .github/workflows/korext.yml:

name: Korext Enforcement
on: [push, pull_request]

permissions:
  contents: read
  security-events: write

jobs:
  enforce:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: Korext/enforce-action@v3
        with:
          api-token: ${{ secrets.KOREXT_API_TOKEN }}

Korext scans your code on every push and PR using the default web policy pack.

How It Works

Install: The action installs the Korext CLI
Scan: Runs korext enforce against your codebase with the selected policy pack
Report: Generates a SARIF file and uploads it to GitHub Code Scanning
Gate: Fails the workflow if critical or high severity violations are found

Violations appear as annotations directly on the PR diff, powered by GitHub Code Scanning.

Inputs

View the full README on GitHub

Details

Repository: github.com/korext/enforce-action
Category: AI & LLM Tools
Package: korext
Transport: STDIO, HTTP
Auth: Required
Registry: MCP registry, active
Tools verified: Nov 6, 2026

Related AI & LLM Tools MCP servers

View all

glim.shglim-sh/glim-mcpGive your agent live data from Twitter, Reddit, the web and GitHub. No API keys, no scraping stack.4stars

Daimonosbeardfaceguy/daimonosAgent-optimized MCP server with structured JSON tools. 20-45% token savings.3stars

Gradus Notationdelmas41/gradusnotationGradus Notation + Harmonic Analyzer + Engraver: music tools for any AI agent. Free, no auth.3stars

Dagpipedevilsfave/dagpipeCrash-proof LLM pipelines. Resumes from failure without a database.3stars

Evc Team Relay Mcpentire-vc/evc-team-relay-mcpMCP server for AI agent read/write access to Obsidian vaults via EVC Team Relay.3stars

Dynamoigetdynamoi/dynamoiPromote a song on Spotify or promote a video on YouTube from your AI agent. Free Smart Links.3stars