Kite LogikDocumentation

OPA/Rego policy enforcement at the agent action layer. Deterministic. Auditable. Open source.

Govern actions, not text

Kite Logik intercepts every tool call, sub-agent spawn, delegation, plan, and resource request. The model can propose anything; only authorised actions execute.

Policy-as-code with OPA/Rego

The same engine that runs Kubernetes admission control and cloud IAM, applied to AI agent actions. Reuse the skills and tooling your platform team already has.

Deterministic and auditable

Same input + same policy = same decision. Every governed event is recorded with the policy version that decided it. Reproducible in tests, defensible in audits.

One pipeline, every framework

11 framework adapters (OpenAI, Agents SDK, LangChain, LangGraph, CrewAI, ADK, Pydantic AI, LlamaIndex, Semantic Kernel, Haystack, Dify) sharing one policy engine and audit log.