Command Palette

Search for a command to run...

UnylyUnyly
Browse all

Clarx

FreeNot checked

Scores agent manifests, validates repo configuration, and generates CI workflows for AI-ready codebases.

GitHubEmbed

About

Scores agent manifests, validates repo configuration, and generates CI workflows for AI-ready codebases.

README

AI-first codebase standard, analysis engine, and CLI.

What's in here

Package Name Description
packages/engine @clarxai/engine Codebase analysis engine — scores repos against the Clarx standard
packages/cli @clarxai/cli CLI tool: clarx score, clarx init, clarx explain
apps/docs Documentation site (Next.js 15 + fumadocs)
standard/ Versioned Clarx AI-First Standard spec and rubric
packages/ui Internal reference components (not published)

Prerequisites

  • Node.js 20+
  • pnpm 10+ (npm install -g pnpm)

Install

git clone https://github.com/clarxai/clarx
cd clarx
pnpm install

Development

Run everything in watch mode:

pnpm dev

Run a specific workspace:

pnpm --filter @clarxai/cli dev        # CLI in watch mode
pnpm --filter @clarxai/engine build   # build the engine
pnpm --filter docs dev                # docs site at localhost:3000

Build

pnpm build

Lint / typecheck

pnpm lint
pnpm typecheck

Format

pnpm format

CLI (clarx)

Install globally (after building)

cd packages/cli
pnpm build
npm link

Commands

clarx score [path] [options]    Score a codebase against the AI-First Standard
clarx init [path]               Generate a starter clarx-manifest.json
clarx explain <rule-id>         Explain a specific rule (e.g. clarx explain C2)

Score options

Flag Description
--format text|json|markdown Output format (default: text)
--min-score <n> Exit 1 if overall score is below n
--min-pillar-score <n> Exit 1 if any pillar score is below n
--ignore <globs> Comma-separated glob patterns to exclude
--verbose Include passing rules in output

Examples

clarx score                                        # score current directory
clarx score ./my-repo --min-score 70               # fail CI if score < 70
clarx score --format json > report.json            # machine-readable output
clarx init                                         # scaffold clarx-manifest.json
clarx explain O1                                   # explain rule O1

Monorepo layout

packages/engine/src/      — analysis engine
packages/cli/src/         — CLI
apps/docs/                — documentation site
apps/docs/content/docs/   — MDX page content
standard/                 — AI-First Standard spec (v0.1)
standard/v0.1.md          — full standard document
standard/rubric/          — scoring rubric
journal/                  — architecture decisions and audits
clarx-manifest.json       — machine-readable repo guidance
packages/ui/src/          — internal reference components (not published)

Contributing

Add a doc page

  1. Create MDX file in apps/docs/content/docs/
  2. Add it to apps/docs/content/docs/meta.json

Add an engine rule

  1. Add rule evaluation to packages/engine/src/scoring/rules.ts
  2. Update pillar score calculation in packages/engine/src/scoring/overall.ts if needed
  3. Add rule explanation to packages/cli/src/commands/explain.ts
  4. Document in apps/docs/content/docs/standard/pillars.mdx

Add a CLI command

  1. Create packages/cli/src/commands/[name].ts
  2. Wire it up in packages/cli/src/cli.ts
  3. Document in apps/docs/content/docs/cli/commands.mdx

from github.com/gernika-labs/clarx

Installing Clarx

This server has no published package — it is built from source. Open the repository and follow its README.

▸ github.com/gernika-labs/clarx

FAQ

Is Clarx MCP free?

Yes, Clarx MCP is free — one-click install via Unyly at no cost.

Does Clarx need an API key?

No, Clarx runs without API keys or environment variables.

Is Clarx hosted or self-hosted?

Self-hosted: the server runs locally on your machine via the install command above.

How do I install Clarx in Claude Desktop, Claude Code or Cursor?

Open Clarx on unyly.org, pick your client tab (Claude Desktop, Claude Code, Cursor) and press Install — the config is generated automatically, no JSON editing.

Related MCPs

Compare Clarx with

Not sure what to pick?

Find your stack in 60 seconds

Author?

Embed badge for your README

Browse similar

All ai MCPs