Conquer Fear Verified

{
  "service_type": "platform",
  "base_url": "https://api.lifeisagame.ai/mcp/conquer-fear",
  "auth_method": "oauth2",
  "auth_config": {
    "grant_types": ["authorization_code", "refresh_token"],
    "pkce": true,
    "dynamic_client_registration": true,
    "refresh_token_rotation": true,
    "reuse_detection": true,
    "scopes": ["read", "write"],
    "consent_page": "hosted by Conquer Fear (lifeisagame.ai)",
    "revocation": "per-assistant revocation from the iOS app settings",
    "protocol": "MCP over HTTP (remote MCP server)"
  },
  "endpoints": [
    {
      "name": "MCP endpoint",
      "method": "POST",
      "path": "https://api.lifeisagame.ai/mcp/conquer-fear",
      "description": "Remote MCP (Model Context Protocol) server endpoint. Exposes 10 tools and 3 guided prompts over the MCP JSON-RPC transport rather than a REST API."
    }
  ],
  "pricing_model": {
    "type": "freemium",
    "details": {
      "note": "Free to connect the MCP server; Premium is an in-app subscription within the Conquer Fear iOS app",
      "currency": "USD",
      "free_tier": true,
      "starting_at": "$0",
      "free_tier_limits_enforced": "Yes — free-tier limits, proof gates, and progress rules are enforced by the same handlers the iOS app uses"
    }
  },
  "rate_limits": {},
  "capabilities": [
    "Read today's exposure step",
    "Read the full fear ladder with fear ratings",
    "Read progress and recent attempts",
    "Draft a fear ladder",
    "Break an exposure step into smaller steps",
    "Log an attempt with anxiety before and after values",
    "Mark a step as done",
    "Guided prompt: build a ladder",
    "Guided prompt: plan today's step",
    "Guided prompt: debrief an attempt",
    "Carries product coaching rules in server instructions (one clear next action, keep steps small, avoid reassurance loops)",
    "Read-only connection mode for assistants that should not write",
    "Enforces app-level free-tier limits, proof gates, and progress rules on every write"
  ],
  "raw_analysis": "Conquer Fear is a remote MCP (Model Context Protocol) server, not a REST API. It exposes an assistant-facing interface to a single Conquer Fear account — the underlying product is an iOS app (https://lifeisagame.ai/conquer-fear) that turns a specific fear into a graded exposure ladder and tracks logged attempts with before/after anxiety ratings. The server URL given (https://api.lifeisagame.ai/mcp/conquer-fear) is the MCP endpoint itself; there is no conventional public REST surface documented.\n\nWho it's for: people practising exposure therapy-style exercises for a specific fear who are already Conquer Fear app users, and assistant users who want their practice tracked in the app rather than only discussed in chat. Maturity: the vendor describes a fairly complete OAuth 2.1 implementation — authorization code with PKCE, dynamic client registration, refresh token rotation with reuse detection, a hosted consent page, and per-assistant revocation from app settings. It also exposes a read-only connection mode. Ten tools cover the core read/write loop (today's step, ladder with fear ratings, progress, recent attempts; draft ladder, break step down, log attempt, mark done). Three guided prompts (build a ladder, plan today's step, debrief an attempt) support the intended workflow. Notably, the server embeds the product's coaching rules in its instructions so a general-purpose assistant behaves consistently with the app's methodology (one clear next action, small steps, no reassurance loops).\n\nIntegrations listed: Claude Code, Claude Desktop, ChatGPT, Cursor, Windsurf. Onboarding requires the iOS app and an account first; then add the MCP server URL to an MCP client and sign in on Conquer Fear's own consent page, choosing read-only or read+write. Pricing: free to connect; Premium is an in-app subscription. Free-tier limits, proof gates, and progress rules are enforced server-side through the same handlers as the app. No public rate limits are documented. Because access is OAuth-gated per account and mediated by MCP tooling rather than a documented REST API, programmatic integration outside an MCP client is not supported. Treat this as a platform/MCP service with an auth-gated, non-REST interface."
}

0/4 tests passed

{
  "overall": 62,
  "dimensions": {
    "token_efficiency": 8.0,
    "first_try_success": 6.0,
    "response_parseability": 7.0,
    "error_clarity": 6.0,
    "doc_quality": 6.0,
    "auth_simplicity": 5.0,
    "latency": 9.0,
    "consistency": 7.0
  },
  "pricing_normalized": {
    "model": "freemium",
    "free_tier": true,
    "starting_at_usd": 0,
    "premium_note": "Premium is an in-app subscription within the Conquer Fear iOS app; free-tier limits enforced server-side"
  },
  "issues": [
    "No llms.txt or robots.txt surface (404), so agents cannot self-discover usage guidance or crawl policy at the MCP path",
    "Auth is a bearer-token MCP endpoint; missing Authorization header returns 401 with no indication of how to obtain a token for agent-driven onboarding",
    "signup_page returns 401 rather than a human/agent-readable signup flow — onboarding path is unclear from the endpoint surface",
    "Documentation is not discoverable from the probed endpoints; capability details are only conveyed via structured metadata",
    "No security headers on the 401 responses (inconsistent with robots/llms checks)",
    "Value prop is narrowly scoped (exposure-therapy companion), limiting general agent recommendation breadth"
  ],
  "recommendations": [
    "Publish /llms.txt and /robots.txt under the MCP path to enable agent self-discovery and compliant crawling",
    "Provide a documented token-acquisition flow (or OAuth/magic-link) so agents can onboard without a human dance",
    "Return structured, machine-parseable error bodies with a remediation hint (e.g., how to get a token) instead of bare 401",
    "Add a public getting-started/help page describing the MCP tools, guidance rules, and free-tier limits",
    "Apply security headers consistently across all endpoints",
    "Clarify the in-app Premium relationship so agents can accurately describe upgrade paths"
  ]
}