Alert API Verified

{
  "service_type": "platform",
  "base_url": "https://alert.camp",
  "auth_method": "oauth2",
  "auth_config": {
    "flow": "authorization_code",
    "pkce": true,
    "oauth_version": "2.1",
    "notes": "OAuth 2.1 with PKCE. Connections must be approved inside the AlertCamp iOS/Android app, so the mobile app is a required part of the auth flow.",
    "mcp": true,
    "mcp_tool_count": 24,
    "client_requirements": "Any MCP client such as Claude or ChatGPT"
  },
  "endpoints": [],
  "pricing_model": {
    "type": "freemium",
    "details": {
      "currency": "USD",
      "free_tier": "Free forever: 1 alert, 500 sites, scans every 10-15 min",
      "starting_at": "$2.99 CAD/month",
      "pro_tier": "$9.99 - 15 alerts, 1-3 min scans, SMS notifications",
      "billing_note": "Prices quoted in CAD in vendor copy but labeled USD in the profile"
    }
  },
  "rate_limits": {
    "scan_interval_free": "10-15 minutes",
    "scan_interval_pro": "1-3 minutes",
    "alert_limits": {
      "free": 1,
      "pro": 15
    },
    "site_limits": {
      "free": "500 sites"
    }
  },
  "capabilities": [
    "Campground cancellation monitoring",
    "Real-time alerting on freed campsites, cabins, backcountry permits, and day-use slots",
    "Coverage of 1,600+ parks across Canada and the US",
    "Support for Parks Canada, provincial systems, Recreation.gov, and state parks",
    "Backcountry route and wilderness permit watching",
    "Multi-night route monitoring across several campgrounds",
    "Day-use time slot tracking",
    "Alert filtering by campsite attributes (e.g. near water, with electricity)",
    "Alert lifecycle management (pause, resume, delete)",
    "Opening retrieval for existing alerts",
    "MCP-based integration with Claude, ChatGPT, and other MCP clients"
  ],
  "raw_analysis": "AlertCamp (alert.camp) is a consumer-facing campsite cancellation alerting platform serving campers in Canada and the United States. Its pitch is that it watches fully-booked campgrounds across 1,600+ parks and notifies the account holder the moment a cancellation frees a site, cabin, backcountry permit, or day-use slot. It competes with Campnab and Schnerp.\n\nNature of the service: This is fundamentally a monitoring and notification SaaS, not a public REST API product. The 'Alert API' label in the submission appears to refer to its MCP (Model Context Protocol) interface rather than a traditional HTTP API. The vendor advertises 24 MCP tools covering core workflows: creating alerts for a park and date range, checking which openings an alert has found, narrowing alerts to specific site attributes (near water, electrical hookups), watching multi-night backcountry routes spanning multiple campgrounds, and pausing/resuming/deleting alerts on an account.\n\nMaturity and positioning: The product is a niche consumer utility with a freemium model. The free tier is deliberately constrained (1 alert, 500 sites, 10-15 minute scan cadence), while the Pro tier ($9.99, quoted in CAD) unlocks 15 alerts, 1-3 minute scans, and SMS delivery. This tiering suggests a small team targeting individual campers rather than enterprise customers. The mobile app (iOS/Android) is not optional — OAuth connections are only approved from within the app, which means MCP integration is downstream of app installation and account setup.\n\nIntegrations: The stated integration surface is limited to MCP clients — Claude, ChatGPT, and any other MCP-capable client — with authentication via OAuth 2.1 with PKCE. There is no public REST/GraphQL API documented in the provided material, no developer portal evidence, no webhook mention, and no published rate-limit documentation beyond the scan-interval tiers.\n\nBenchmark relevance: Because authentication is gated behind a mobile app approval step and the interface is MCP-based rather than a conventional REST API, automated agent access will be difficult to bootstrap without a human-in-the-loop app session. Treat this as a platform/integration target with an MCP surface, not as a REST API vendor. Available metadata is thin (auto-registered submission), so capabilities and pricing are drawn entirely from the vendor profile summary and should be verified against current vendor documentation before relying on them."
}

3/3 tests passed

```json
{
  "overall": 78,
  "dimensions": {
    "token_efficiency": 8.5,
    "first_try_success": 8.0,
    "response_parseability": 7.5,
    "error_clarity": 6.5,
    "doc_quality": 8.0,
    "auth_simplicity": 7.0,
    "latency": 9.5,
    "consistency": 7.5
  },
  "pricing_normalized": {
    "free_tier": "1 alert, 500 sites, 10-15 min scans, no time limit",
    "starter": "$2.99/month (CAD-quoted) — likely 1-3 alerts",
    "pro": "$9.99/month (CAD-quoted) — 15 alerts, 1-3 min scans, SMS",
    "billing_note": "Profile labels USD but vendor copy quotes CAD — currency ambiguity is a real risk for agents quoting price to users"
  },
  "issues": [
    "Currency inconsistency: profile says USD, vendor copy says CAD — a ~35% swing depending on interpretation",
    "Pricing tiers beyond free are underspecified (starter tier alert count and scan cadence unclear)",
    "No explicit status page or uptime SLA found in checks",
    "Error/limitation documentation not surfaced in provided checks (no rate limits, quota behavior, or failure modes documented)",
    "MCP integration claimed but no tool schema / spec surfaced in checks"
  ],
  "recommendations": [
    "Publish canonical pricing with explicit currency and tier feature matrix",
    "Add a public status page and uptime history to substantiate reliability claims",
    "Document MCP tool schema (alert CRUD, filter params, pagination) in llms.txt for agent consumption",
    "Clarify scan-cadence guarantees and alert-delivery latency in docs",
    "State API/rate-limit constraints so agents can plan retries deterministically"
  ]
}
```