Language
The Language section covers the tools used to build, orchestrate, and deploy autonomous agents within the Paranet.
The Paranet supports two primary ways to define actor behavior: Paraflow (the native orchestration language) and the Python Agent SDK (the integration bridge). Understanding when to use each is critical for building scalable, performant autonomous systems.
Paraflow: The Core Orchestrator
Paraflow is a domain-specific language purpose-built for Stateful Orchestration. It is designed for "Goal-Seeking" agents where persistence and deterministic logic are paramount.
- Deterministic Logic: Paraflow focuses strictly on monitoring, planning, and orchestrating activities.
- Durable Workflows: Paraflow natively supports "Goal Trees" (Rules vs. Tasks). Rules allow for complex branching and sub-goal resolution that persists across system restarts.
- State Management: Best-in-class for actors that must "remember" state, manage long-running loops, or synchronize data across a fleet (e.g., a "Fleet Sync" actor).
Python Agent SDK: The Integration Bridge
The Python Agent SDK allows developers to build actors using pure Python. It acts as an abstraction layer, generating the necessary Paraflow code behind the scenes. It is the preferred choice for Execution and Integration.
- Service Integration: Seamlessly bridge the Paranet with ecosystems like ROS, OpenCV, or NVIDIA Isaac Sim.
- Data Manipulation: Ideal for heavy computation or data-heavy sidecars where Python’s library ecosystem (NumPy, Pandas) excels.
- Ephemeral Tasks: Python is optimized for "Tasks" (execution) rather than complex "Rules" (logic branching).
Choosing Your Tool
| Feature | Paraflow (Native) | Python SDK |
|---|---|---|
| Primary Strength | Orchestration & State | Integration & Computation |
| Workflow Depth | Multi-level Goal Trees (Rules) | Flat Workflows (Tasks) |
| Data Persistence | Native Ledger / State DB | Ephemeral (External DB required) |
| Learning Curve | Moderate (DSL-specific) | Low (Standard Python) |
| Best For | "The Brain" (Decision Making) | "The Muscles" (Action/Calculation) |