Etch

About

Every decision made by an AI agent—including prompts, tool calls, edits, and multi-agent handoffs—is captured in a secure, verifiable history that cannot be altered without detection. Easily trace the reasoning behind any action without searching through text logs. Provide a single link as definitive proof for customers or auditors inquiring about your agent’s activity on any given day. This system integrates seamlessly beneath your existing agent stack or context graph. It is compatible with Claude Code, Cursor, Codex, Continue, and any tool capable of sending webhook events.