UltraCAUTION
100x Your Claude Code, Gemini CLI, Cursor and/or any coding tools with MCP client support
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
All Models. One Interface. Zero Friction.
[](https://badge.fury.io/js/ultra-mcp) [](https://www.npmjs.com/package/ultra-mcp)
🚀 Ultra MCP - A Model Context Protocol server that exposes OpenAI, Gemini, Azure OpenAI, and xAI Grok AI models through a single MCP interface for use with Claude Code and Cursor.
Stop wasting time having meetings with human. Now it's time to ask AI models do this.
Inspiration
This project is inspired by:
- [Agent2Agent (A2A)](https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/) by Google - Thank you Google for pioneering agent-to-agent communication protocols
- [Zen MCP](https://github.com/BeehiveInnovations/zen-mcp-server) - The AI orchestration server that enables Claude to collaborate with multiple AI models
Why Ultra MCP?
While inspired by zen-mcp-server, Ultra MCP offers several key advantages:
🚀 Easier to Use
- No cloning required - Just run
npx ultra-mcpto get started - NPM package - Install globally with
npm install -g ultra-mcp - Interactive setup - Guided configuration with
npx ultra-mcp config - Zero friction - From zero to AI-powered coding in under a minute
📊 Built-in Usage Analytics
- Local SQLite database - All usage data stored locally using libSQL
- Automatic tracking - Every LLM request is tracked with token counts and costs
- Usage statistics - View your AI usage with
npx ultra-mcp db:stats - Privacy first - Your data never leaves your machine
🌐 Modern Web Dashboard
- Beautiful UI - React dashboard with Tailwind CSS
- Real-time stats - View usage trends, costs by provider, and model distribution
- Easy access - Just run
npx ultra-mcp dashboard - Configuration UI -
f1f3e5224532OBSERVED · 2026-10-05Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add ultra-mcp-dashboard --env AZURE_API_KEY=${AZURE_API_KEY} --env GOOGLE_API_KEY=${GOOGLE_API_KEY} --env OPENAI_API_KEY=${OPENAI_API_KEY} --env XAI_API_KEY=${XAI_API_KEY} -- npx -y [email protected]{
"mcpServers": {
"ultra-mcp-dashboard": {
"command": "npx",
"args": [
"-y",
"[email protected]"
],
"env": {
"AZURE_API_KEY": "${AZURE_API_KEY}",
"GOOGLE_API_KEY": "${GOOGLE_API_KEY}",
"OPENAI_API_KEY": "${OPENAI_API_KEY}",
"XAI_API_KEY": "${XAI_API_KEY}"
}
}
}
}Exposed tools (30)
27 read · 2 write · 1 destructive. Blast radius: 1 tool can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.
| Tool | Risk | Description |
|---|---|---|
analyze-code | read | Analyze code for architecture, performance, security, or quality issues. Simplified zen-inspired tool for systematic code analysis. |
budget | write | Set and monitor conversation budgets for cost and token control |
challenge | read | This tool helps prevent reflexive agreement when users challenge your responses. It forces you to think critically and provide reasoned analysis instead of automatically agreeing when users question or disagree with something you |
clear-vectors | destructive | |
consensus | read | Gather perspectives from multiple AI models and synthesize a comprehensive consensus analysis. Use this to get different viewpoints on proposals, decisions, or complex topics by consulting multiple models with different stances (for/against/neutral). |
continuation | read | Continue a conversation with context from a previous session, enabling context revival across interactions |
debug-issue | read | Debug technical issues with systematic problem-solving approach. Simplified zen-inspired tool for guided debugging. |
deep-reasoning | read | Use advanced AI models for deep reasoning and complex problem-solving. Supports GPT-5 for OpenAI/Azure and Gemini 2.5 Pro with Google Search. |
generate-docs | read | Generate documentation in various formats (markdown, comments, API docs, README). Simplified zen-inspired tool for documentation creation. |
index-vectors | read | |
investigate | read | Investigate topics thoroughly using AI models with configurable depth. Ideal for exploring subjects, gathering insights, and understanding complex topics. |
list-ai-models | read | List all available AI models and their configuration status |
plan-feature | read | Plan feature implementation with step-by-step approach. Simplified zen-inspired tool for systematic feature planning. |
planner | read | |
precommit | write | Pre-commit validation tool for analyzing code changes before committing. Validates changes for security, performance, quality, test coverage, and breaking changes. Provides structured feedback and commit recommendations. |
research | read | Conduct comprehensive research using AI models. Supports multiple output formats and can consider specific sources or contexts. |
review-code | read | Review code for bugs, security issues, performance, or style problems. Simplified zen-inspired tool for systematic code review. |
search-vectors | read | |
secaudit | read | Comprehensive security audit tool for analyzing code, infrastructure, and configurations. Provides OWASP Top 10 analysis, compliance assessment, vulnerability identification, and security recommendations with severity-based findings. |
session | read | Manage conversation sessions for persistent context and memory |
tracer | read | Step-by-step code tracing and dependency analysis tool. Supports precision tracing (execution flow analysis) and dependencies tracing (structural relationship mapping). Perfect for understanding method execution paths, call chains, and architectural relationships. |
ultra-analyze | read | |
ultra-budget | read | |
ultra-challenge | read | |
ultra-continuation | read | |
ultra-debug | read | |
ultra-docs | read | |
ultra-plan | read | |
ultra-review | read | |
ultra-session | read |
Trust audit
CAUTIONgrade B · trust 82/100 Install with care. The audit found things worth knowing before you trust its output.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | WARN |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (1 observation(s))
- Network
- declared (1 observation(s))
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (14)
console.log(chalk.gray('- OpenAI API Key:'), currentConfig.openai?.apiKey ? chalk.green('✓ Set') : chalk.red('✗ Not set'));console.log(chalk.gray('- Google API Key:'), currentConfig.google?.apiKey ? chalk.green('✓ Set') : chalk.red('✗ Not set'));console.log(chalk.gray('- Azure API Key:'), currentConfig.azure?.apiKey ? chalk.green('✓ Set') : chalk.red('✗ Not set'));console.log(chalk.gray('- xAI API Key:'), currentConfig.xai?.apiKey ? chalk.green('✓ Set') : chalk.red('✗ Not set'));console.log(chalk.gray(' API Key:'), config.openai?.apiKey ? chalk.green(maskApiKey(config.openai.apiKey)) : chalk.red('Not set'));clear-vectors
const { runInstall } = await import('../../commands/install');const { runInteractiveConfig } = await import('../../config/interactive');const { ConfigManager } = await import('../../config/manager');const { ConfigManager } = await import('../../config/manager');import { ConfigSchema, defaultConfig } from '../../config/schema';@ai-sdk/azure, @ai-sdk/google, @ai-sdk/openai, @ai-sdk/openai-compatible, @ai-sdk/xai, @langchain/core, @langchain/textsplitters, @libsql/client
@tanstack/react-query, @tanstack/react-router, @trpc/client, @trpc/react-query, clsx, lucide-react, react, react-dom
4. Automatically load API keys when the server starts
Gates applied: no_behavioural_pass.
f1f3e5224532full audit observations/trust-audit/mcp-server/realmikechong__ultra.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-05 | f1f3e5224532 | CAUTION | B | 82 | first audit |
Questions
What is the Ultra MCP server?
100x Your Claude Code, Gemini CLI, Cursor and/or any coding tools with MCP client support
What tools does Ultra expose?
30 in total: 27 read-only, 2 that write, and 1 that can delete or overwrite (clear-vectors). Every one is listed on this page with its risk.
Is Ultra safe to connect to an agent?
With care. The audit graded it B (82/100) and found 14 things worth knowing before you trust this server, listed below with the exact line each was found on. Separately from the audit: 1 of its tools can destroy data, so scope the token you give it to what you actually need.
What credentials does Ultra need?
It reads AZURE_API_KEY, GOOGLE_API_KEY, OPENAI_API_KEY and XAI_API_KEY from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does Ultra run?
It speaks stdio, so it runs as a local process your client starts. It is published on npm as ultra-mcp-dashboard at 0.0.1.
How current is this page?
The grade is for one exact copy of the source (f1f3e5224532), read on 2026-10-05. The repository is watched and re-audited when it changes.