# zero-core-intel

**Category:** 💰 Finance & Fintech  
**Repository:** https://github.com/meltingpixelsai/harvey-intel  
**GitHub Stars:** 1  
**npm Downloads (last month):** 291  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/zero-core-intel

## Description
DrainBrain token safety, CORTEX trading signals, social trends. Zero Core Intel.

## Tools
Capabilities this server exposes over MCP:

- **list_tools** — List all tools with pricing
- **health** — Server status and payment config
- **scan_token_preview** — Quick risk level (LOW/MEDIUM/HIGH/CRITICAL)
- **scan_token** — Full DrainBrain risk analysis - outcome-calibrated score, honeypot, rug stage, risk flags, temporal prediction (beta)

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "zero-core-intel": {
    "command": "npx",
    "args": ["-y","mcpay"]
  }
}
```

## Documentation

## What zero-core-intel does

zero-core-intel provided AI agents with token-safety data for Solana assets through RugSlayer’s DrainBrain service. Its purpose was to help an agent inspect a token’s risk before making a decision, with both a fast categorical preview and a more detailed analysis.

The project was formerly named Harvey Intel. It was built by RugSlayer and categorized as a finance and fintech tool because its primary function was crypto-token risk analysis rather than portfolio management or trading execution.

## How it works

Agents connected to a remote MCP endpoint at `https://agents.rugslayer.com/mcp`. The server exposed free tools for discovering available operations and checking service status. A preview request returned one of four risk levels: LOW, MEDIUM, HIGH, or CRITICAL.

The detailed `scan_token` operation used DrainBrain to produce an outcome-calibrated risk score, honeypot information, rug-stage classification, risk flags, and a beta temporal prediction. That operation was priced at $0.01 and used USDC payments on Solana when accessed through the x402 flow.

Two access models were described. API-key users sent a bearer token in the MCP connection’s `Authorization` header. The documented tiers included a free allowance of 100 calls per day, a Pro plan with unlimited calls, and a PAYG plan with a monthly fee plus usage charges. Alternatively, MCPay could intercept a paid-tool request, arrange a USDC payment on Solana, obtain verification from the PayAI facilitator, and forward the request.

## Setup and configuration

The documented remote configuration used an MCP server entry with the endpoint URL. API-key configurations added an `Authorization: Bearer ...` header. Direct connections without credentials were intended for free tools only.

The x402 example used `npx mcpay connect` with a Solana secret key and the Solana network selected. That workflow required the agent to respond to a `PAYMENT_REQUIRED` result before the paid tool could run.

Claude Code could be configured with the HTTP transport using:

```bash
claude mcp add harvey-intel --transport http https://agents.rugslayer.com/mcp
```

## Tools and capabilities

- `list_tools`: returned the available MCP tools and their prices.
- `health`: reported server status and payment configuration.
- `scan_token_preview`: supplied a quick LOW, MEDIUM, HIGH, or CRITICAL risk classification.
- `scan_token`: returned the full DrainBrain analysis, including score, honeypot result, rug stage, risk flags, and beta temporal prediction.

## Limitations and notes

The service was retired on October 23, 2026. The remote endpoint now responds with HTTP 410, and the repository is archived for reference rather than operation. RugSlayer stated that the DrainBrain API and its payment routes end on October 29, 2026, so API keys, payment configuration, and the documented live setup should not be used for a new deployment.

The detailed scan was marked as a paid operation and its temporal prediction was identified as beta. DrainBrain scores were calibrated against the real outcomes of tokens previously scanned by RugSlayer, making the output an analytical signal rather than a guarantee about future token behavior.

_Full upstream README: https://allmcps.com/mcp/zero-core-intel/readme_