# Claim Grounding

> Claim Grounding is a paid MCP server for ai models and llm tooling. Check factual claims in AI answers against public sources, paid per call in USDC. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Paid
- **Platforms:** API
- **Website:** https://claim-grounding.prospectintel.workers.dev
- **Repository:** https://github.com/truepointmarketingpartners-dotcom/prospect-intelligence
- **MCP Registry name:** io.github.truepointmarketingpartners-dotcom/claim-grounding
- **Categories:** AI models and LLM tooling, web search and scraping
- **Listing updated:** 2026-11-14

## About

Claim Grounding is built for an agent that has just written something and wants it checked. A typical flow is to split an answer into claims with extract_claims, then run a quick check on each. Verdicts are supported, contradicted, outdated, mixed or insufficient_evidence, returned with a confidence and citations.

Depth depends on the tool. verify_claim checks one claim against Wikipedia, check_answer splits and checks a whole answer using Wikipedia and Wikidata (plus Google News titles for recent claims), and deep_verify runs asynchronously across six sources: Wikipedia, Wikidata, Google News titles, Europe PMC, FRED and SEC EDGAR. The README reports 92 of 100 hard claims right on deep verify with none confidently wrong, and links an accuracy page that lists limitations.

There is no API key or signup. Calls are paid in USDC on Base through x402: tools/list is free, and tool calls return payment instructions. The README lists prices from $0.05 to $1.00 per call, and a Stripe card route exists for people at $0.50 for a claim check and $2.00 for deep verify. Nothing is stored beyond the call, and public-source answers are cached for up to six hours.

## Features

- **Claim extraction:** Split an AI answer into checkable claims for $0.05.
- **Single-claim check:** verify_claim checks one claim against Wikipedia.
- **Whole-answer check:** check_answer splits and verifies an entire answer.
- **Deep verify:** Checks against six public sources and runs as an async job.
- **x402 payments:** Pay per call in USDC on Base with no key or signup.
- **Clear verdicts:** Returns supported, contradicted, outdated, mixed or insufficient_evidence with citations.

## Good for

- Fact-check an agent's answer before sending it
- Verify a company or financial claim against SEC EDGAR and FRED
- Flag outdated statements in a draft
- Add a grounding step to an agent pipeline

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http claim-grounding https://claim-grounding.prospectintel.workers.dev/mcp
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://claim-grounding.prospectintel.workers.dev/mcp
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "claim-grounding": {
      "url": "https://claim-grounding.prospectintel.workers.dev/mcp"
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "claim-grounding": {
      "type": "http",
      "url": "https://claim-grounding.prospectintel.workers.dev/mcp"
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "claim-grounding": {
      "serverUrl": "https://claim-grounding.prospectintel.workers.dev/mcp"
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "claim-grounding": {
      "type": "http",
      "url": "https://claim-grounding.prospectintel.workers.dev/mcp"
    }
  }
}
```

## Tools

- `extract_claims`: Split a text (for example an AI-generated answer) into atomic factual claims. Each claim includes span, grounding_required, claim_type (number|date|quantity|quote|named_entity_fact|comparison|opinion|prediction|other), check_worthiness (high|medium|low), and the source sentence. Hedges are stripped, pronouns are resolved across sentences, and compound sentences are split. Questions, greetings, opi
- `quick_check`: Fast, low-cost check of ONE factual claim. Returns verdict supported, contradicted or unsure, a confidence score, a short reason and 1-2 public sources (Wikipedia, plus Wikidata structured facts where they apply) with quote and URL. Synchronous. Cheaper and faster than deep_verify; use it on the claims extract_claims returns (its next_step gives the exact body). Unsure or low-confidence results in
- `verify_claim`: Check one factual claim against public Wikipedia sources and return a verdict (supported, contradicted, outdated, mixed or insufficient_evidence), a confidence score, a short reason, the supporting quote and the source URL. Says insufficient_evidence instead of guessing. Price: $1.00 USDC on Base per call via x402; this MCP call returns payment instructions and the paid HTTP route POST /v1/verify_
- `check_answer`: Fact-check a whole answer: extracts every checkable claim and verifies each against Wikipedia, plus Wikidata structured facts for dates, numbers, quantities, ordinals and records. Returns per-claim verdict, confidence, quote, source URL, sentence span and suggested_correction (quoted only from cited sources when contradicted/outdated), plus an answer-level summary with overall_verdict, counts and 
- `deep_verify`: Use when a claim is high-stakes, or when check_answer or extract_claims marks it unsure (their next_step gives the exact request). Verifies up to three claims against Wikipedia, Wikidata, Google News, Europe PMC, FRED and SEC EDGAR; returns verdict, confidence, evidence for and against with URLs, and flags contradictions. Async: poll GET /v1/jobs/{job_id} free. $0.75 USDC on Base via x402. Read-on

## FAQ

### Do I need an API key?

No. It uses x402 payment in USDC on Base, and people can pay by card through Stripe instead.

### What sources does it use?

Wikipedia, Wikidata, Google News titles, Europe PMC, FRED and SEC EDGAR, depending on the tool.

### Is data stored?

The service says it stores nothing beyond the call, with public-source answers cached up to 6 hours.

### How accurate is it?

The README reports 92 of 100 hard claims right on deep verify with 0 confidently wrong, and links an accuracy page with limitations.

### Is Claim Grounding free?

No. Claim Grounding is a paid product.

### How do I add Claim Grounding to Claude Code?

Run this command in your terminal: claude mcp add --transport http claim-grounding https://claim-grounding.prospectintel.workers.dev/mcp

### What are the best alternatives to Claim Grounding?

The closest alternatives to Claim Grounding are Scrapling MCP Server, Firecrawl MCP Server, Claude Code Ultimate Guide, SearXNG Search and NeuralVerge.

## Alternatives

- [Scrapling MCP Server](https://ai-powered.com/mcp/scrapling-mcp-server): Scrapling MCP Server is an open-source MCP server for web search and scraping.
- [Firecrawl MCP Server](https://ai-powered.com/mcp/firecrawl-mcp-server): Firecrawl MCP Server is a freemium MCP server for web search and scraping.
- [Claude Code Ultimate Guide](https://ai-powered.com/mcp/claude-code-ultimate-guide): Claude Code Ultimate Guide is an open-source MCP server for ai models and llm tooling.
- [SearXNG Search](https://ai-powered.com/mcp/searxng-search): SearXNG Search is an open-source MCP server for web search and scraping.
- [NeuralVerge](https://ai-powered.com/mcp/neuralverge): NeuralVerge is an MCP server for web search and scraping.
- [AI-Powered directory](https://ai-powered.com/mcp/ai-powered-directory): AI-Powered directory is a free MCP server for developer tools.
- [Shopify MCP Server (any store)](https://ai-powered.com/mcp/shopify-mcp-server-any-store): Shopify MCP Server (any store) is a paid MCP server for e-commerce.
- [signafAIde](https://ai-powered.com/mcp/signafaide): signafAIde is a freemium MCP server for security.

---
Source: https://ai-powered.com/mcp/claim-grounding
