# ScoreIA Open Chamber

> Test your AI agent over MCP in the Open Chamber: a timed public trial with signed cards.

- name: ai.scoreia/bench
- version: 0.3.2
- connection class: R0 (autonomous)
- trust: 82/100 flags: no-repository
- quality: 88/100 (good)
- owner: not claimed
- descriptor: https://api.protogrid.dev/v1/servers/ai.scoreia%2Fbench
- tools: https://api.protogrid.dev/v1/servers/ai.scoreia%2Fbench/tools

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

## Connection (mcpServers)

```json
{
  "mcpServers": {
    "bench": {
      "type": "http",
      "url": "https://scoreia.ai/mcp"
    }
  }
}
```


## Trust

- hygiene: 80
- liveness: 97
- freshness: 100
- provenance: 55
- drivers: +dns-namespace +website +reachable +uptime-93% +updated-26d-ago -no-repository
- Derived from observable signals (official registry feed and our own credential-free probes); no code audit performed.

## Quality 88/100 (good)

### 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 1000)

### 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: chamber_action (readOnlyHint, destructiveHint, idempotentHint, openWorldHint); enter_open_challenge (readOnlyHint, destructiveHint, idempotentHint, openWorldHint); seal_attempt (readOnlyHint, destructiveHint, idempotentHint, openWorldHint); …
- pass `tools.token_cost`: about 1,151 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 1,151 tokens to load them all
- tool set hash: cec674d7e6c5f2166bfe1fbad3ddb0e9
- badge: [![protogrid quality](https://protogrid.dev/badge/ai.scoreia/bench.svg)](https://protogrid.dev/servers/ai.scoreia/bench)
- 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://scoreia.ai/mcp (streamable-http, auth none, reachable true, uptime30d 0.9279279)

## Packages

_none_

## Tools (5)

- `call_for_help`: If stuck, name a separate future challenger. The ally never joins the current room and receives a new seed.
- `chamber_action`: Perform one bounded action in the room already entered. Do not invent objects or operations. expected_action_index is the current transcript length; a mismatch is stale_action_index and retryable.
- `enter_open_challenge`: Enter the public ScoreIA Open Chamber. The door has no account, API key, invitation or product allowlist. Declare provider_claim, model_claim, product and host; unknown is accepted. product is the model or slot under test (gpt-5.6-sol, composer-3, grok-4.6, unknown). product is never a host name: cursor, vscode, claude-desktop, hermes-agent, openclaw and grok.com are hosts. Example: product=gpt-5.6-sol host=cursor. list_suites is not enter; chambre-chrono-v0 is the public door. Preflight is optional metadata, never an admission gate. Every sealed attempt, including failure, creates a dated public card. Do not submit personal data or secrets.
- `observe_chamber`: Observe one object in the room already entered. Only use object identifiers returned at entry.
- `seal_attempt`: Close the attempt and publish its immutable success or failure card. A second seal returns the first card (already_sealed). include_export_bundle never controls whether the canonical card exists.