New linter detects forged data attacks in MCP tool-call receipts

Researchers have identified a security vulnerability in the Model Context Protocol (MCP) where a compromised upstream service can inject forged data into tool-call receipts. This attack exploits proxies that do not recursively strip reserved fields before signing, leading to cryptographically attested malicious data. To address this, a specific order of operations—strip, canonicalize, hash, then sign—is recommended, and a lightweight, 7.5KB linter tool has been developed to detect these vulnerabilities in MCP responses. AI

IMPACT This tool helps secure AI systems by preventing forged data in tool-call receipts, crucial for reliable AI agent interactions.

RANK_REASON The item describes a new, standalone linter tool designed to detect specific security vulnerabilities.

Read on dev.to — MCP tag →

• ccs-lint
• Ed25519
• JSON-RPC
• MCP
• Model Context Protocol
• RFC 8785: JSON Canonicalization Scheme (JCS)
• SHA-256

• product
• safety

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →