# Norma by Quality Clouds [Health: Active]

**Category:** 💻 Developer Tools  
**Repository:** https://github.com/qualityclouds/norma-mcp  
**GitHub Stars:** 5  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/norma-by-quality-clouds

## Description
Deterministic AI code review, with an audit record. Governance inside the agent loop.

## Claude Desktop Quick Installation
Remote MCP endpoint (confidence: high). Install path detected from listing signals. Add as a URL/SSE server in your client:

```json
"mcpServers": {
  "norma-by-quality-clouds": {
    "url": "https://api.qualityclouds.ai/mcp"
  }
}
```

## Documentation

## What Norma by Quality Clouds MCP server does

Norma by Quality Clouds MCP server places stack-aware coding rules inside an agent's development workflow. It detects the technologies used by a linked repository and activates relevant rulesets across security, performance, scalability, manageability, maintainability, and architecture.

The service is intended for agents that need more than an informal review. It returns rule metadata and fix guidance, evaluates individual files, exposes unresolved findings from the most recent full repository scan, and records actions taken during a coding task. The resulting record can distinguish rules verified as compliant, violations fixed with file and line details, and violations avoided while code was generated.

Rules are available for multiple languages, frameworks, and platforms. The README specifically lists TypeScript, JavaScript, Python, PHP, Node, React, Vite, Supabase, FastAPI, SQLAlchemy, and Adobe Magento, among others.

## How it works

The server is hosted at `https://api.qualityclouds.ai/mcp` and uses Streamable HTTP. Authentication uses OAuth, with browser sign-in on the first connection. A repository must be linked before the other tools are used in a workspace.

A typical coding session follows this order:

1. Link the current workspace with `link_repository`.
2. Call `get_rulesets` to identify the repository's stack.
3. Retrieve each applicable ruleset with `get_rules_for_ruleset`.
4. Write or change code using those rules as context.
5. Run `live_check` on every created or modified file.
6. Call `register_applied_actions` with the exact rule IDs and recorded outcomes.

For existing problems, an agent can call `get_open_issues` and work from the findings produced by the repository's last full scan. A live file check does not store the file, while action registration contributes to the compliance audit trail.

## Setup and configuration

Add the remote endpoint to an MCP-compatible client and complete the OAuth sign-in flow in a browser. Claude Code can register it with `claude mcp add --scope user --transport http norma https://api.qualityclouds.ai/mcp`. Cursor can use the same URL in `.cursor/mcp.json` under an MCP server named `norma`.

Lovable supports Norma as a custom MCP connector using HTTP and OAuth. For audit registration in Lovable, provide the repository's public GitHub URL in chat so the server can link the repository; the README notes that Lovable's tokenized Git remote is not sent to third parties. The service has a permanent free tier, and the README does not specify environment variables or local installation steps.

## Tools and capabilities

- `link_repository` associates the current workspace with Norma and is required once before other calls.
- `get_rulesets` detects the repository technology stack without configuration.
- `get_rules_for_ruleset` returns rule IDs, names, severities, impact areas, descriptions, and remediation guidance.
- `live_check` evaluates one file against the organization's rules without storing the file.
- `get_open_issues` returns unresolved findings from the latest full repository scan.
- `register_applied_actions` records compliance checks, fixes, prevented violations, affected files and lines, and the model involved.

## Limitations and notes

Norma by Quality Clouds MCP server requires an MCP client that supports its remote HTTP connection and OAuth flow. The first workspace connection must link the repository, and the agent should fetch rules before making changes. The README states that live checks can work in Lovable before linking, but audit registration will fail until a public GitHub repository URL is supplied.

The server's rules are selected from detected stack technologies, so the applicable coverage depends on what the repository scan identifies. The README does not provide a local server mode, environment-variable configuration, or a license declaration. Repository findings and the Production-Ready Score are available in the linked Quality Clouds workspace.

_Full upstream README: https://allmcps.com/mcp/norma-by-quality-clouds/readme_