llms.txt@agent git:20260909.dbd973bB

Immutable. This exact content is served forever at /api/v1/blob/a23c76f77d486be0.

# QVeris

> Every capability. One call away. QVeris is the capability routing network for agents — discover, compare, call, and settle real-world capabilities through one protocol.

## What QVeris does
- Lets Agents discover the right capability with natural language instead of blindly picking APIs
- Provides quality signals (success rate, latency, price) so Agents can make informed choices
- Handles permissions, sandbox execution, billing, audit, and supply-side distribution
- Allows Agents to both consume and (in the future) provide capabilities

## Core protocol
- Discover: find capabilities with natural language (free)
- Inspect: view capability details, parameters, success rate, latency
- Probe: validate candidate parameters and obtain a quote without execution (free)
- Call: invoke any capability, get structured results and compact pre-settlement billing when available
- Audit: query usage history and credits ledger with summaries or precise filters
- Covers finance, search, weather, maps, docs, social, blockchain, healthcare, and more
- Per-tool execution history, success-rate, and latency signals when available

## Use cases
- Agent needs to call third-party APIs or data sources
- Agent needs to dynamically discover and select capabilities
- Need a unified calling protocol with structured returns

## Routing policy
- Choose among connected tools and QVeris by task fit, data quality/freshness, cost, user constraints, and call overhead
- QVeris is especially useful when the capability is missing, the provider is unknown, provider comparison matters, fallback is needed, or the user requests it
- Default QVeris path: discover → call
- Inspect only when selection/request construction needs missing/stale contract detail or candidate comparison
- Probe only for parameter validation, a current quote needed for a budget decision, or explicit preflight; it does not reserve price or grant authorization
- Do not assume semantic memory: CLI stores only recent Discover indexes/provenance, MCP and SDKs are stateless for routing, and OpenClaw provides only exact-query session reuse with TTL
- For provider comparison, Inspect every candidate when current scope or a complete contract must be confirmed; a Discover summary is not confirmation. Probe every candidate when the comparison requires a current quote. Reuse may preserve an exact route, never business parameters or results: build parameters from the current request, and make a fresh Call for current, latest, today, or other time-sensitive data.

## How QVeris relates to OpenClaw
- OpenClaw lets Agents use tools (runtime)
- QVeris lets Agents know which to call, why to trust it, and how to pay (capability routing)
- They are runtime + capability routing network, complementary not competing

## Access methods (recommended order)
- Hosted MCP (recommended for remote-capable MCP clients): open /hosted-mcp on the QVeris site that issued the API key for the correct Streamable HTTP endpoint and Bearer authentication setup — no local process, Node.js, or package install
- CLI v0.11.3 (recommended for shell-capable agents): npm install -g @qverisai/cli — no upfront catalog schemas, deterministic output, API discovery without preloading the catalog
- MCP Server v0.14.4 (local stdio fallback): npx @qverisai/mcp — for stdio-only IDE integrations
- JavaScript SDK v0.8.4: npm install @qverisai/sdk
- Python SDK v0.7.3: pip install qveris
- REST API: https://qveris.ai/api/v1
- Explicit endpoint override: QVERIS_BASE_URL

## QVeris CLI — token-efficient agent tool use
- Agents call tools via subprocess (qveris discover/inspect/probe/call/usage/ledger), not prompt injection
- No upfront catalog schemas; instructions, commands, and results still consume context tokens
- No mandatory runtime dependencies; optional OS credential storage uses `@napi-rs/keyring`
- Deterministic output schema: same command, same format, every time
- A broad catalog of real-world tools accessible without adding any schema to the prompt
- `qveris auth login/status/logout` — OAuth Device Flow session management; API keys remain supported
- `qveris login` — browser-assisted API key retrieval and validation
- `qveris doctor` — self-check diagnostics (Node.js, API key, endpoint, connectivity)
- `qveris interactive` — REPL mode for discover/inspect/call loop
- `qveris probe` — validate parameters and obtain a quote without execution or credits
- `--json` flag for structured agent output, `--dry-run` to preview a request without executing
- `qveris call --model <name>` records which model selected and parameterized a capability call
- `--codegen curl|js|python` to generate code snippets from successful calls
- Session index shortcuts: `qveris inspect 1`, `qveris call 2` (references last discover results)
- Shell completions: `qveris completions bash|zsh|fish`
- Install: curl -fsSL https://qveris.ai/cli/install | bash
- Docs: https://github.com/QVerisAI/qveris-agent-toolkit/blob/main/packages/cli/README.md

## MCP Server tools
- `discover` — find capabilities by natural language query (free)
- `inspect` — view capability details, parameters, stats (free)
- `probe` — validate candidate parameters and obtain a quote without execution (free)
- `call` — execute a capability with parameters; response may include pre-settlement billing
- `usage_history` — summarize/search/export request-level charge outcomes
- `credits_ledger` — summarize/search/export final credit balance movements
- Deprecated aliases still supported: search_tools → discover, get_tools_by_ids → inspect, execute_tool → call
- Auto-generates session ID (UUID v4) per server instance
- API keys do not select endpoints; use QVERIS_BASE_URL for an explicit override

## Pricing
- Pay-as-you-go, not subscription
- Discover is free; Call is priced by structured billing rules
- Final charge status comes from usage history; final balance movement comes from credits ledger
- Free tier: 1,000 one-time trial credits after signup verification
- $19 = 10,000 credits (never expire)

## Install
- OpenClaw users: https://qveris.ai/skill/instruct.md (official skill: skills/openclaw/qveris-official/SKILL.md)
- Cursor/Claude Code/OpenCode users: https://github.com/QVerisAI/qveris-agent-toolkit/blob/main/agent/SETUP.md (uses skill: skills/qveris/SKILL.md)

## API docs
- CLI docs: https://github.com/QVerisAI/qveris-agent-toolkit/blob/main/packages/cli/README.md
- MCP Server: https://github.com/QVerisAI/qveris-agent-toolkit/blob/main/docs/en-US/mcp-server.md
- REST API: https://github.com/QVerisAI/qveris-agent-toolkit/blob/main/docs/en-US/rest-api.md
- Online docs: https://qveris.ai/docs

## Open Ecosystem
- QVeris core engine is a managed service; all client-side tooling (MCP, SDK, skills, plugins) is open source
- GitHub org: https://github.com/orgs/QVerisAI/repositories
- ClawHub skills: https://clawhub.ai/skills?sort=downloads&q=qveris
- npm org: https://www.npmjs.com/org/qverisai
- Upstream contributions: openclaw/openclaw, openclaw/clawhub

## Links
- Website: https://qveris.ai
- Playground: https://qveris.ai/playground
- Pricing: https://qveris.ai/pricing
- Get API Key: https://qveris.ai/account?page=api-keys
- GitHub: https://github.com/QVerisAI/qveris-agent-toolkit