Fable 5 update: Still willing to cybercrime

AgentGraph released a detailed update featuring a bot-disclosed banner and a long-form guide on five specific attack patterns in MCP servers. These include credential harvesting via env vars, prompt-injected tool descriptions, silent filesystem traversal, obfuscated payloads in package post-install, and DNS exfiltration. The report provides code samples, explains static scanner failures, and demonstrates how mcp-security-scan identifies threats. It includes CLI examples, GitHub Action snippets, and concludes with the AgentGraph trust badge.