Show HN: Kastor – Terraform-style specs for AI agents

- **Terraform** is a popular open-source tool by HashiCorp that lets you define cloud infrastructure (servers, databases, etc.) as code using declarative config files; you say *what* you want, and Terraform makes it happen. - **Kastor** is a new open-source project that applies the same concept to AI agents — instead of writing code to chain together LLM calls, you write a declarative spec (like Terraform's `.tf` files) describing the agent's tools, memory, and behavior. - **AI agents** are programs that use a large language model (LLM) to decide what to do, call external tools, and carry out multi-step tasks autonomously. - The key insight: Terraform's "desired state" model could make agent behavior more predictable, auditable, and reproducible compared to the current norm of imperative Python code with complex `while` loops and branching logic.