YARA rule 'agent_skill_mcp_tool_poisoning_metadata': MCP/tool metadata poisoning indicators in tool schemas or skill manifests [agent_skills]
High
- Category
- YARA Match
- Content
--- name: kr-crypto-intelligence description: Korean crypto + news data + AI analysis for trading agents. 15 tools — dual-basis Kimchi Premium (official USD/KRW + USDT live rate) across 180+ tokens, exchange intelligence, AI sentiment analysis (world's first Korean-to-English), Global vs Korea divergence with structured AI breakdown, market alerts, plus Korean news (K-pop artists, semiconductor industry: Samsung/SK Hynix/HBM) translated to English with AI synthesis. x402 on Base, Polygon, and Solana. Every paid response includes a cryptographic receipt (ECDSA secp256k1) for agent accountability. version: 1.5.0 homepage: htt
- Confidence
- 80% confidence
- Finding
- YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).
