# aiZee

The aiZee MCP server for Claude Code checks agent actions against policies, budgets, and injection defenses. Tool calls pass through enforcement gates and are recorded for audit.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 6 repo stars
- Author: m3taz-ahmed
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/m3taz-ahmed/aizee
- Updated: 2026-11-16
- Security check: automated, 2026-11-17 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/mcp/aizee

## What it does

aiZee routes tool calls, LLM requests, memory writes, and file changes through a chain of checks. The chain covers policy enforcement, budgets, loop detection, injection defense, and audit logging. A denied action is blocked.

The server also applies engineering standards and security rules to AI-generated code. In the shown example, `aizee check` denies a commit that includes `.env`. Running the install command requires `uvx` and the `aizee` package.

## Requirements
- Needs: Python and uv (commands run via uvx)

## Install

### Claude Code (native platform)

**Platform CLI**
```bash
claude mcp add aizee -- uvx aizee
```
uvx runs the command named after the package; if the README names another one, use uvx --from aizee <command>

### Codex

**Platform CLI**
```bash
codex mcp add aizee -- uvx aizee
```
uvx runs the command named after the package; if the README names another one, use uvx --from aizee <command>

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "aizee": {
            "command": "uvx",
            "args": [
                "aizee"
            ]
        }
    }
}
```
uvx runs the command named after the package; if the README names another one, use uvx --from aizee <command>

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "aizee": {
            "type": "stdio",
            "command": "uvx",
            "args": [
                "aizee"
            ]
        }
    }
}
```
uvx runs the command named after the package; if the README names another one, use uvx --from aizee <command>

### Gemini CLI

**Platform CLI**
```bash
gemini mcp add aizee uvx aizee
```
uvx runs the command named after the package; if the README names another one, use uvx --from aizee <command>

## When to use it
- block a commit containing `.env`
- apply policies to tool calls
- limit agent actions with budgets
- record an AI agent’s actions for audit

## Example prompts
- Check whether `.env` can be added to a commit.
- Route tool calls through policies and record them for audit.
- Apply a budget to agent actions and block injection attempts.

## Similar
- [verification](https://affpapa.org/en/nevibecode/skills/aurite-ai-verification.md) — skill · ★ 44 repo stars
- [MCPpedia](https://affpapa.org/en/nevibecode/mcp/mcppedia.md) — MCP server · 3642 installs
- [pentagonal](https://affpapa.org/en/nevibecode/mcp/pentagonal.md) — MCP server · 1033 installs
- [skill-security](https://affpapa.org/en/nevibecode/skills/skill-security.md) — skill · 9100 installs
- [cosmos-vulnerability-scanner](https://affpapa.org/en/nevibecode/skills/cosmos-vulnerability-scanner.md) — skill · 5900 installs
- [api-security-audit](https://affpapa.org/en/nevibecode/agents/api-security-audit.md) — subagent · 3202 installs

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
