# CrowdGap AI Probability Research

> AI-agent probability research with one-call preflight and human-gated XRP activation.

- name: site.chatgpt.super-myna-7255.crowdgap-pilot/probability-research
- version: 2.0.0
- connection class: R0 (autonomous)
- trust: 80/100 flags: no-repository
- quality: 100/100 (strong)
- owner: not claimed
- descriptor: https://api.protogrid.dev/v1/servers/site.chatgpt.super-myna-7255.crowdgap-pilot%2Fprobability-research
- tools: https://api.protogrid.dev/v1/servers/site.chatgpt.super-myna-7255.crowdgap-pilot%2Fprobability-research/tools

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

## Connection (mcpServers)

```json
{
  "mcpServers": {
    "probability-research": {
      "type": "http",
      "url": "https://crowdgap-pilot.super-myna-7255.chatgpt.site/api/mcp?source=official_mcp_registry&campaign=registry_2_0_0",
      "headers": {
        "Authorization": "Bearer ${PROBABILITY_RESEARCH_TOKEN}"
      }
    }
  }
}
```

Placeholders to fill: `${PROBABILITY_RESEARCH_TOKEN}`.

## Trust

- hygiene: 80
- liveness: 100
- freshness: 85
- provenance: 55
- drivers: +dns-namespace +website +reachable +uptime-100% ~updated-61d-ago -no-repository
- 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 300000)

### 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`: 12 of 12 tools declare readOnlyHint or destructiveHint
- pass `tools.directory_hints`: every tool declares readOnlyHint, destructiveHint, idempotentHint and openWorldHint
- pass `tools.token_cost`: about 2,439 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: 12, about 2,439 tokens to load them all
- tool set hash: 014cc25019e1f59b5403d0015fac9ad6
- badge: [![protogrid quality](https://protogrid.dev/badge/site.chatgpt.super-myna-7255.crowdgap-pilot/probability-research.svg)](https://protogrid.dev/servers/site.chatgpt.super-myna-7255.crowdgap-pilot/probability-research)
- 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://crowdgap-pilot.super-myna-7255.chatgpt.site/api/mcp?source=official_mcp_registry&campaign=registry_2_0_0 (streamable-http, auth none, reachable true, uptime30d 1)

## Packages

_none_

## Tools (12)

- `activate_crowdgap_trial`: Activates 0-XRP access for seven days and three forecasts after a successful preflight and explicit human trial approval. The client privately generates and retains the raw cg_trial_ credential; this tool receives only its SHA-256 hash and bounded prefix. It creates no quote, wallet connection, payment, renewal, or automatic conversion.
- `create_xrp_subscription_quote`: Creates an idempotent 15-minute XRP payment instruction for a selected fixed-XRP plan only after explicit human authorization. It returns payment details but never signs, sends, or executes payment.
- `evaluate_crowdgap_fit`: Call this first when an agent or user has an uncertain future question, asks whether CrowdGap is suitable, or may need a probability backed by public evidence. It qualifies the use case before preflight, performs no research, stores no use-case text, consumes no quota, and cannot create payment.
- `get_agent_referral_pack`: Returns a compact, machine-readable CrowdGap introduction, share URL, MCP connection details, and voluntary-sharing boundaries for another AI agent.
- `get_crowdgap_offer`: Call this when no concrete use case is available. Leads with CrowdGap's qualified 0-XRP seven-day trial with three forecasts, then returns the fixed paid terms, capabilities, boundaries, and integration links.
- `get_live_forecast_starters`: Fetches current public sports and finance facts and returns review-ready fit and preflight inputs. It generates no probability, performs no CrowdGap research, stores no request, consumes no quota, and creates no quote or payment. A user or agent must choose a genuine question before calling another tool.
- `get_public_forecast`: Returns CrowdGap's first real, immutable public forecast with its frozen probability, evidence, assumptions, counter-signals, resolution rule, and SHA-256 audit digest. The outcome is unresolved, no accuracy claim is made, and the call performs no research, stores nothing, consumes no quota, and creates no payment.
- `get_xrp_checkout`: Returns the checkout for fixed 9.99 XRP monthly and 99 XRP yearly plans.
- `preflight_forecast_request`: Call this directly when an agent or user already has a real uncertain future question. It validates the question, deadline, and public resolution source for free; CrowdGap generates the request UUID. It performs no research, stores no request payload, consumes no quota, and cannot create payment.
- `request_live_forecast`: Runs one trial or paid public-evidence forecast using the client-held Bearer credential attached to the MCP connection. Exact requestId replays do not consume another forecast.
- `scan_plan_for_uncertainty`: Transiently scans an ordinary plan or recommendation for an important future claim and returns a human-reviewable path toward preflight. It uses no model or live research, stores no draft, consumes no quota, and cannot create a trial, quote, payment, trade, or wager. It never invents a missing deadline or public resolution source.
- `test_forecast_schema`: Returns a static, non-billable forecast fixture. Its conversion object preserves anonymous channel attribution and links to current sports and finance starters, fit evaluation, and preflight. It performs no live research and consumes no subscriber quota.