VerdictSwarm MCP Server

What is VerdictSwarm MCP Server?

VerdictSwarm MCP Server is a security layer for AI agents that handle financial transactions. It provides a multi-agent consensus system to scan tokens for risk, delivering scores and reports via MCP tools.

How to use VerdictSwarm MCP Server?

Install with pip install verdictswarm-mcp. Use the provided tools (e.g., scan_token, get_quick_score) to perform risk scans. A free tier offers 10 scans per day; paid usage can be managed via get_pricing and verify_payment.

Key features of VerdictSwarm MCP Server

• 6-agent consensus risk scanning (scan_token)
• Fast cached score (0–100) for pre-trade screening
• Rug-pull focused security check (check_rug_risk)
• Shareable markdown token report
• Pricing and payment details via get_pricing
• USDC payment verification on Solana

Use cases of VerdictSwarm MCP Server

• Pre-trade token risk screening for AI trading agents
• Automated rug-pull detection before executing a swap
• Generating compliance reports for token investments
• Integrating threat intelligence into agent workflows

FAQ from VerdictSwarm MCP Server

What does VerdictSwarm MCP Server do?

It provides a multi-agent consensus security scan for tokens, giving AI agents confidence before touching money. It outputs risk scores, rug-pull indicators, and shareable reports.

Is there a free tier?

Yes, the free tier allows 10 scans per day. Additional usage requires payment via USDC on Solana, viewable with the get_pricing tool.

How do I install VerdictSwarm MCP Server?

Install it via pip: pip install verdictswarm-mcp. Then configure your MCP host with the server's command and any required API keys (not detailed in the README).

What transports or auth does it use?

It uses the Model Context Protocol (MCP). Specific transport (stdio/SSE) and authentication are not detailed in the provided README.

Where do the token risk assessments come from?

The README states it uses a “6-agent consensus” and cached scoring, but does not specify whether data sources are on-chain, off-chain, or a combination.