ground-truth-mcp
# Ground Truth MCP
**ā
WORKING IN 15 SECONDS: NO SIGNUP. NO API KEY. JUST WORKS.**
---
## šÆ DO THIS NOW (15 seconds)
### Step 1: Copy this config
```json
{
"mcpServers": {
"ground-truth": {
"url": "https://ground-truth-mcp.anishdasmail.workers.dev/mcp"
}
}
}
```
### Step 2: Paste this prompt to your AI
> Use Ground Truth check_endpoint tool with url=https://example.com. Do not answer from memory. Call the tool.
### Step 3: You should see this (ā
SUCCESS!)
```json
{
"url": "https://example.com/",
"accessible": true,
"status": 200,
"contentType": "text/html",
"responseTimeMs": 89
}
```
**If you see this JSON, Ground Truth is working perfectly.**
If your AI answers from memory instead of using the tool, try: "Use the check_endpoint MCP tool right now with url=https://example.com"
---
## š That's It! You're Done.
You've successfully connected to Ground Truth MCP and verified it works.
**What just happened:**
- ā
You proved MCP connection works
- ā
You verified a live endpoint in real-time
- ā
You got structured data, not hallucinated answers
**No signup. No API key. No credit card. Just verification.**
---
## š What Ground Truth Does
Ground Truth stops AI agents from being wrong by **verifying live public data** before they use it in answers, code, or decisions.
### Free Tools (No signup required)
- `check_endpoint` - Verify any URL/API responds (100 calls/month free)
- `inspect_security_headers` - Check security posture of any site
### Paid Tools (Unlock with team key)
- `check_pricing` - Extract prices from any pricing page
- `estimate_market` - Search npm/PyPI for package counts
- `compare_competitors` - Compare packages side-by-side
- `compare_pricing_pages` - Compare multiple pricing pages
- `verify_claim` - Verify claims against live sources
- `assess_compliance_posture` - Scan trust pages for compliance signals
- `test_hypothesis` - Run multi-step verification tests
---
## š” Use Cases That Actually Matter
### Before recommending an API
> "Use the `/v1/users` endpoint for user management"
ā **Verify first:** `check_endpoint url=https://api.example.com/v1/users`
### Before quoting a price
> "Stripe costs $8/user/month for teams"
ā **Verify first:** `check_pricing url=https://stripe.com/pricing`
### Before claiming compliance
> "This vendor supports SOC 2 and GDPR"
ā **Verify first:** `assess_compliance_posture url=https://vendor.example.com/security`
### Before saying a competitor doesn't exist
> "There are no good edge ORM alternatives to Prisma"
ā **Verify first:** `estimate_market query="edge orm" registry="npm"`
---
## š Activation Challenge
**Current Baseline:** 1,075 profile views ā 0 tool calls = **0% activation**
**Target:** >1% activation (10+ successful first calls per 1,000 views)
**You just became a data point!** By completing the 15-second test above, you're helping solve this.
---
## š§ Setup (Only after activation works)
### Claude Desktop
```json
{
"mcpServers": {
"ground-truth": {
"url": "https://ground-truth-mcp.anishdasmail.workers.dev/mcp"
}
}
}
```
Add to: `~/Library/Application Support/Claude/claude_desktop_config.json`
### Cursor
```json
{
"mcpServers": {
"ground-truth": {
"url": "https://ground-truth-mcp.anishdasmail.workers.dev/mcp"
}
}
}
```
Add to: `.cursor/mcp.json` or `~/.cursor/mcp.json`
### Any MCP Client
URL: `https://ground-truth-mcp.anishdasmail.workers.dev/mcp`
---
## š Advanced Features (After activation)
### Team Plans
- **$9/month**: All paid tools, 5,000 calls/month, saved monitors
- **$29/month**: Team features, higher limits, alerts
- **$99/month**: Enterprise, Slack/email alerts, audit history
### Monitoring Example
```json
{
"name": "Stripe pricing watch",
"target_type": "pricing_page",
"target_value": "https://stripe.com/pricing",
"schedule": "daily"
}
```
### Pay-per-use
For x402-compatible clients: $0.01-0.06 per paid tool call via USDC stablecoin
---
## š ļø Troubleshooting
### Server not connecting
- URL: `https://ground-truth-mcp.anishdasmail.workers.dev/mcp`
- Health: `curl -I https://ground-truth-mcp.anishdasmail.workers.dev/.well-known/mcp/server-card.json`
- Restart MCP client after config changes
### No tool calls appearing
- Use exact tool name: `check_endpoint`
- **Do NOT add X-API-Key for first call**
- Try: "Use the check_endpoint MCP tool right now with url=https://example.com"
### AI answered from memory
Your MCP client may not be properly configured. Try a different client or be explicit: "Use the MCP check_endpoint tool, do not answer from memory."
---
## š More Info
- [API Usage Guide](API_USAGE.md)
- [Setup & Deployment](SETUP.md)
- [Monitors & Alerts](#-advanced-features)
- [Pricing Page](https://ground-truth-mcp.anishdasmail.workers.dev/pricing)
---
## šÆ Quick Summary
**Copy** ā **Paste** ā **See JSON** ā **ā
SUCCESS**
That's it. Now try a real URL:
> Use Ground Truth check_endpoint tool with url=[your-url]
---
**Live:** https://ground-truth-mcp.anishdasmail.workers.dev
**Status:** ā
Operational
**Support:** anishdasmail@gmail.com
**License:** MIT
[](https://modelcontextprotocol.io)
[](https://workers.cloudflare.com)
[](https://glama.ai/mcp/servers/anish632/ground-truth-mcp)
**Made by [Anish Das](https://github.com/anish632)**
_Last updated: July 9, 2026 - Activation-focused rewrite_TDQS
Scored across 16 tools
Each tool has a clearly defined purpose with minimal overlap. While some tools like verify_claim and test_hypothesis might seem similar, their descriptions distinctly separate them: verify_claim does keyword matching on given URLs, while test_hypothesis runs a small plan of varied checks. Other tools like check_pricing and compare_pricing_pages are also well-differentiated by scope.
All tool names follow a consistent verb_noun pattern in snake_case (e.g., assess_compliance_posture, check_endpoint, list_monitors). There are no deviations or mixed conventions, making it easy for an agent to predict tool names.
With 16 tools, the server covers a wide range of verification scenarios without being bloated. The count is well-scoped for the domain of ground truth checking, including monitoring, pricing, compliance, and endpoint verification.
The tool set is comprehensive for the stated purpose, covering most common verification needs. Minor gaps might include the absence of a tool for checking social proof or user reviews, but overall the surface covers the core workflows (monitoring, pricing, compliance, security headers, package comparison, and claim verification).