# PubFi MCP

> Use MCP 2026-07-28 discovery or 2025-11-25 initialize to list and execute PubFi routes.

- name: ai.pubfi/mcp
- version: 0.2.84
- connection class: R0 (autonomous)
- trust: 91/100 flags: multi-version-spam
- quality: 100/100 (strong)
- owner: not claimed
- descriptor: https://api.protogrid.dev/v1/servers/ai.pubfi%2Fmcp
- tools: https://api.protogrid.dev/v1/servers/ai.pubfi%2Fmcp/tools

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

## Connection (mcpServers)

```json
{
  "mcpServers": {
    "pubfi": {
      "type": "http",
      "url": "https://mcp.pubfi.ai",
      "headers": {
        "Authorization": "Bearer ${PUBFI_TOKEN}"
      }
    }
  }
}
```

Placeholders to fill: `${PUBFI_TOKEN}`.

## Trust

- hygiene: 75
- liveness: 100
- freshness: 100
- provenance: 85
- drivers: +repository +dns-namespace +website +reachable +uptime-100% +updated-8d-ago -multi-version-spam(100)
- Derived from observable signals (official registry feed and our own credential-free probes); no code audit performed.

## Quality 100/100 (strong)

### protocol: 100 (weight 25)

- pass `protocol.modern`: supports 2026-07-28 (server/discover)
- pass `protocol.stateless`: answers without a session
- pass `protocol.transport`: streamable HTTP
- pass `protocol.list_ttl`: tool list is cacheable (ttlMs 30000)

### 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: 100 (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
- pass `tools.annotations`: 4 of 4 tools declare readOnlyHint or destructiveHint
- pass `tools.directory_hints`: every tool declares readOnlyHint, destructiveHint, idempotentHint and openWorldHint
- pass `tools.token_cost`: about 1,090 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: 4, about 1,090 tokens to load them all
- tool set hash: 7f4863f11754660582ddd993430ca8ab
- badge: [![protogrid quality](https://protogrid.dev/badge/ai.pubfi/mcp.svg)](https://protogrid.dev/servers/ai.pubfi/mcp)
- 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.pubfi.ai (streamable-http, auth none, reachable true, uptime30d 1)

## Packages

_none_

## Tools (4)

- `pubfi.capabilities.get`: Return the full typed request, response, metering, and readiness contract for one exact capability id from the installed Registry generation.
- `pubfi.capabilities.list`: Enumerate the complete installed Registry v2 catalog as deterministic compact pages. The server does not rank, infer intent, or choose a capability.
- `pubfi.route.execute`: Execute one exact GET or POST path through the authenticated PubFi account lane and the same Registry v2 matcher and typed executor as the HTTP gateway. Append :free only when the capability catalog publishes that exact account-scoped free variant.
- `pubfi.substrate.runtime_upgrade.verify`: Verify one reviewed Substrate System.apply_authorized_upgrade extrinsic through the authenticated account lane. Supply the expected Blake2b-256 hash of the FRAME System AuthorizedUpgrade code payload, not the active :code hash at the apply block. The server streams and discards provider runtime code, persists only a compact proof, and returns independent comparison evidence.
