Last released Feb 22, 2026
SIGIL security layer for AI agent tool calls — scans MCP tool arguments for leaked secrets, enforces policies, and writes audit logs
Supported by