# sh.clearhouse/clearhouse

> One search across every agent registry: 100K+ MCP/A2A/x402 services, signed receipts.

- name: sh.clearhouse/clearhouse
- version: 0.8.0
- connection class: R0 (autonomous)
- trust: 90/100
- quality: 83/100 (good)
- owner: not claimed
- descriptor: https://api.protogrid.dev/v1/servers/sh.clearhouse%2Fclearhouse
- tools: https://api.protogrid.dev/v1/servers/sh.clearhouse%2Fclearhouse/tools

**An agent can connect right now, no human step.** Remote endpoint, no authentication, reachable on the last probe.

## Connection (mcpServers)

```json
{
  "mcpServers": {
    "clearhouse": {
      "type": "http",
      "url": "https://mcp.clearhouse.sh/mcp"
    }
  }
}
```


## Trust

- hygiene: 100
- liveness: 100
- freshness: 85
- provenance: 75
- drivers: +repository +dns-namespace +reachable +uptime-100% ~updated-46d-ago
- Derived from observable signals (official registry feed and our own credential-free probes); no code audit performed.

## Quality 83/100 (good)

### protocol: 88 (weight 25)

- pass `protocol.modern`: supports 2026-07-28 (server/discover)
- pass `protocol.stateless`: answers without a session
- pass `protocol.transport`: streamable HTTP
- warn `protocol.list_ttl`: tools/list declares ttlMs 0, so clients re-fetch it on every use

### authorization: n/a (weight 15)

- n/a `auth.prm`, `auth.as_metadata`, `auth.cimd`: no remote uses OAuth
- n/a `auth.secret_in_url`: no templated URL

### tool hygiene: 71 (weight 30)

- pass `tools.descriptions`: every tool has a description
- pass `tools.description_length`: descriptions are concise
- pass `tools.schemas`: every tool has an object input schema
- fail `tools.annotations`: no tool declares readOnlyHint or destructiveHint, so clients cannot tell safe calls from risky ones
- fail `tools.directory_hints`: no tool declares all four of readOnlyHint, destructiveHint, idempotentHint and openWorldHint; missing: index_stats (readOnlyHint, destructiveHint, idempotentHint, openWorldHint); quote_call (readOnlyHint, destructiveHint, idempotentHint, openWorldHint); call_service (readOnlyHint, destructiveHint, idempotentHint, openWorldHint); …
- pass `tools.token_cost`: about 872 tokens to load every tool
- pass `tools.api_dump`: tools are not a one-to-one API dump

### stability: 100 (weight 15)

- pass `stability.changes`: no tool changes in 30 days
- pass `stability.rug_pull`: no tool changed its meaning under the same name

### dependencies: n/a (weight 15)

- n/a `deps.known_vulns`, `deps.mcp_sdk_version`, `deps.resolvable`: no npm or PyPI package

- tools: 5, about 872 tokens to load them all
- tool set hash: be005cd7cc0ef4f96b206da5a698ced7
- badge: [![protogrid quality](https://protogrid.dev/badge/sh.clearhouse/clearhouse.svg)](https://protogrid.dev/servers/sh.clearhouse/clearhouse)
- Checks run on what our credential-free, read-only probes observe; tools are never called and no code audit is performed.

### Tool changes

_No tool definition changes recorded._

## Remotes

- https://mcp.clearhouse.sh/mcp (streamable-http, auth none, reachable true, uptime30d 1)

## Packages

_none_

## Tools (5)

- `call_service`: Call a listed remote service THROUGH Clearhouse and get a signed receipt.

    Routes the call via the matching protocol adapter (MCP / A2A / HTTP-x402),
    enforces spend caps, settles x402 paywalls (dry-run by default), and
    returns the result plus a cryptographically signed receipt binding
    intent -> execution -> settlement -> outcome.

    Args:
        listing_id: id from search_agents results
        tool: downstream tool name (omit to list the service's tools)
        args_json: JSON object of tool arguments
        caller_ref: opaque caller identifier for the receipt
        payment_json: for paid services — JSON {"authorization": typed_data.message,
            "signature": "0x..."} produced by signing quote_call's typed_data with
            your own key. Clearhouse relays it; it never sees your private key.
    
- `index_stats`: How many listings Clearhouse currently indexes, broken down by source registry.
- `quote_call`: Step 1 of a paid call (non-custodial). Fetches the service's payment
    requirements and returns the exact EIP-712 typed data your agent must sign
    with ITS OWN key. Nothing is charged until you sign and pass the result to
    call_service. If the service is free, says so.

    Args:
        listing_id: id from search_agents
        payer_address: your agent's wallet address (the key that will sign)
        args_json: JSON object of query args for the service
    
- `search_agents`: Search all indexed agent registries at once.

    Args:
        query: capability keywords, e.g. "pdf extraction", "solana rpc", "payments"
        remote_only: only return services callable over the network right now
        limit: max results (1-50)
    
- `service_evidence`: Observed facts about a listing: live probe results (handshake, latency,
    HTTP status) and routed-call history (count, success rate, distinct
    callers). Includes an arithmetic summary_index with published methodology
    and a confidence tier. This is EVIDENCE for the caller's own decision —
    not a safety rating, and Clearhouse never selects or executes on your
    behalf.