Klent
Guardrails and observability for AI agents in production.
Klent is a developer SDK and policy engine that adds guardrails and observability to AI agent tool calls. It allows teams to allow, pause, rewrite, or redirect any tool call—such as database writes or API calls—based on configurable YAML policies, with support for human-in-the-loop approval flows. Every decision is logged as a structured, audit-grade event queryable from a dashboard.
Developers integrate a two-line TypeScript or Python SDK that routes every tool call through a synchronous policy engine, which can allow, block, pause for human approval, rewrite inputs, or redirect calls to safer surrogates in under 50ms.
developers and teams running AI agents in production
Background.
- Status
- launched
- Business model
- unknown
Similar projects.
Editorial take on the space this project sits in — momentum signals, adjacent moves, our call on whether the wedge is real. Get pinged when we publish a new read or when the landscape shifts.
Have a take on this space?
Tell us what you’d build differently, where you think the incumbents miss, or what we’ve gotten wrong about this project. Comments + reactions are coming soon.