# QVeris official

The core skill that lets any supported agent discover, inspect, and call real-world capabilities.

## Agent Install Policy

Agents must get explicit user confirmation before installing a skill, writing configuration, or changing the local environment.

1. Read this guide and the manifest: https://qveris.ai/skills/qveris-official/manifest.json
2. Explain to the user what will be installed and which QVeris API actions may be used.
3. Ask for explicit approval before running an install command or changing local configuration.
4. After approval, install the skill and run one starter prompt.

## Official GitHub Source

- Repository: https://github.com/QVerisAI/open-qveris-skills
- Skill path: qveris-official
- Skill source: https://github.com/QVerisAI/open-qveris-skills/tree/main/qveris-official
- Clone and inspect: `git clone https://github.com/QVerisAI/open-qveris-skills.git && cd open-qveris-skills/qveris-official`

## Install Commands

- OpenClaw: `openclaw skills install qveris-official`
- CLI: `npm install -g @qverisai/cli`

## Starter Prompts

### First capability call
A small end-to-end request that proves discover, inspect, and call work.

```text
Discover weather capabilities, inspect the best candidate, and get the current weather in Tokyo. Explain which capability you used and why.
```

### Capability routing test
Ask the agent to find the right provider instead of naming an API manually.

```text
Find a reliable capability for real-time stock quotes, inspect the top options, then call one for AAPL. Include success rate, latency, and estimated cost if available.
```

## QVeris API Actions

- Discover `POST /search`: Find the right real-world capability without manually choosing an API vendor.
- Inspect `POST /tools/by-ids`: Read parameters, examples, quality signals, latency, and billing rules before execution.
- Call `POST /tools/execute`: Execute the chosen capability and return structured results with auditable usage.

## Estimated QVeris Usage

Installation and catalog discovery are free. Discover and Inspect do not consume credits; only successful Call execution follows the selected provider billing rule.

- Typical paid Call executions: 1-3
- Planning estimate: 1-20 credits
- Free actions: Discover, Inspect
- Paid actions: Call
- Note: Use Inspect to view the exact billing_rule before calling. Call responses include pre-settlement billing when available, and final charges are auditable in usage_history and credits_ledger.

## Related Cases

- [QVeris skill for OpenClaw](https://qveris.ai/blog/qveris-skill-for-openclaw) - How to add QVeris as a reusable capability layer inside OpenClaw.
- [Use QVeris in Cursor](https://qveris.ai/blog/qveris-in-cursor) - A step-by-step path for coding agents that need live tools and data.
- [QVeris CLI workflow](https://qveris.ai/blog/qveris-cli) - Use QVeris as a subprocess-friendly capability interface.

## Human Page

https://qveris.ai/skills/qveris-official