# QVeris Agent Toolkit [Health: Active]

**Category:** 🔗 Aggregators  
**Repository:** https://github.com/QVerisAI/qveris-agent-toolkit  
**GitHub Stars:** 264  
**Views:** 2  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/mcp-202

## Description
Discover, inspect and call ranked external data & tool APIs with unified billing and usage audit

## 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": {
  "qveris-agent-toolkit": {
    "url": "https://img.shields.io/npm/v/@qverisai/mcp?label=mcp"
  }
}
```

## Documentation

## What QVeris Agent Toolkit does

QVeris Agent Toolkit exposes QVeris’s professional data and tool marketplace through MCP. An assistant can search for services using a natural-language description, examine a candidate’s supported scope, invoke the selected capability, and review execution or credit outcomes through the available interfaces. The catalog covers areas including finance, search, weather, maps, documents, social, blockchain, and healthcare.

The toolkit is intended as an additional source of tools rather than a replacement for local or native capabilities. Use it when an existing tool is insufficient, a provider must be located at runtime, multiple providers need comparison, or the user specifically requests an external service.

## How it works

The normal workflow starts with `discover`, which returns candidate services and is free. `inspect` can provide additional information about parameters, supported signals, and billing when a contract is incomplete, stale, or needs comparison. `probe` is available when a current quote or parameter validation is needed. After selecting a service, `call` executes it and can return structured pre-settlement billing details.

Usage results can be reviewed through `usage_history`, while `credits_ledger` shows credit-balance movements. Provider metadata may include execution history, success rates, and latency signals. A previously used route can be reused to identify an exact service path, but request parameters and results should be treated as fresh; current or time-sensitive requests require a new call.

## Setup and configuration

The repository provides the MCP server alongside a CLI, JavaScript and Python SDKs, skills, and REST API documentation. The hosted MCP endpoint is `https://mcp.qveris.ai/mcp`. Cursor users can install the repository’s marketplace plugin, enter `QVERIS_API_KEY` in the plugin configuration, start a new agent session, and check that the QVeris tools are available. Claude Code users can follow the repository’s agent setup process. Gemini CLI can install the repository as an extension and store the API key as a sensitive extension setting.

For CLI-based access, install `@qverisai/cli` with npm or the project’s shell installer, then use `qveris discover`, `qveris inspect`, and `qveris call`. Browser OAuth is supported where available; API keys should remain in a private environment or client secret store.

## Tools and capabilities

- Discover services with natural-language requests.
- Inspect service scope, parameters, signals, and billing details.
- Probe candidates for validation or current quotes.
- Call selected external data and tool APIs.
- Review usage history and credit-ledger records.
- Access service signals such as latency and success rate when supplied.

## Limitations and notes

Discovery does not confirm a complete or current provider contract. For comparisons that depend on current scope, inspect each candidate; for quote-based comparisons, probe each one. Calls are subject to the selected capability’s billing rules. A signup trial provides one-time credits after verification, and paid usage is handled through QVeris credits.

_Full upstream README: https://allmcps.com/mcp/mcp-202/readme_