T H E W A L K

Systems Architect, Engineering, Narrative.

← Back to main

Actor-Orchestrator: Agentic Swarm Infrastructure

[ AUTHORIAL INTENT & AI DISCLOSURE ]

This post was drafted with Gemini-3-flash to synthesize technical documentation and system architecture logs for the Actor-Orchestrator project. The final narrative, S5 Protocol definitions, and system diagrams were manually verified for technical accuracy.

Forensic Hygiene Active
View Policy Standard →

Most people see AI as a chatbot. I see it as a distributed systems problem.

Actor-Orchestrator Dashboard

The Actor-Orchestrator is a “Unix-for-AI” architecture designed to manage concurrent, autonomous agent workstreams using gemini-cli and zmx. It moves beyond simple prompt-response loops into a robust, hierarchical system capable of executing complex, long-horizon tasks with high fidelity and total isolation.

The “Unix-for-AI” Philosophy

The core of this system is the belief that AI agents should operate like Unix processes: isolated, specialized, and steerable. By using the Actor Model pattern, we delegate tasks to isolated sub-agents (Workers) that operate in their own persistent terminal sessions. This prevents file-system conflicts, eliminates logical crosstalk, and ensures that every contribution is atomic and verifiable.

The Hierarchy of Intelligence

The system operates through three distinct layers:

  1. Meta-Orchestrator (Teddy): The high-level personal architect that bootstraps the environment and monitors the overall system trajectory.
  2. Supervisor Actor: The project manager that monitors the .tasks/ directory and TODO.md blackboard. It delegates tasks via a deterministic dispatch engine.
  3. Worker Actors: The distributed engineers. Each worker handles a specific task: Clone → Branch → Implement → Verify → Commit → Pull Request.

The S5 Protocol: Atomic Contribution

Every worker follows the S5 Protocol, a strict lifecycle designed for maximum reliability:

  • Isolation: Every task runs in a dedicated branch (task-<ID>-<slug>).
  • Hermetic Execution: Workers run in unique temporary workspaces.
  • Local Verification: No code is committed without automated verification.
  • Automated PRs: Pull requests are created via the GitHub CLI (gh) and reported back to the master Command Center.

Swarm Command Center

Observability is the bridge between autonomy and trust. The Swarm Command Center provides a high-fidelity TUI for real-time monitoring of the active swarm:

  • Active Reasoning: Live previews of worker “thoughts” and pulsating heartbeats.
  • Task Ledger: A sorted history of task states.
  • System Events: A feed of automated PR detections and infrastructure events.

Vision: The Cognitive Multiplier

The Actor-Orchestrator is more than a tool; it’s a Cognitive Multiplier. By offloading high-throughput implementation to parallel agents while maintaining strict architectural oversight, we can reach production-ready states in days, not weeks.

Explore the source on GitHub: ilteris/actor-orchestrator

← Back to main