control-loop

Getting the most out of your hobby servo

Inverted Pendulum in python with pybox2d. Fuzzy and PID simulation included.

Closed-loop graphs for LLMs: generate, judge, retry until a score, then emit a reward for training. Python package (actuate-ai) plus a control-console UI.

A software factory for AI coding agents. Hand it a spec; it plans, builds, measures the result with checks the agent did not write, feeds the gap back, and stops only when independent checks agree. You stay on the loop, not in it.

GitOps continuous-reconciliation engine for Cisco IOS XE. Reads desired state from Git, compares live device state via RESTCONF, writes only the delta via NETCONF/YANG. Idempotent, self-healing, model-driven, a Kubernetes-style control loop for network config

A tiny, dependency-free engine for iterative refinement loops: propose, check, revise, repeat.

The dumb, tireless driver — Box 2 of The Machine. A deterministic, zero-token control loop that drives long-running code work to a verified definition of done. Code-work sibling of Conductor.

SUBIT‑64 is a minimal 6‑bit protocol for deterministic agent control. It defines a compact state model, a strict transition layer, and canonical JSON schemas that make agent behavior transparent, interoperable, and implementation‑agnostic

Adaptive vLLM/GPU serving tuner for Kubernetes: a SelfTune-style control loop that auto-tunes replica count and vLLM --max-num-seqs from live metrics, with Prometheus/Grafana observability and a web UI

Circuit breaker for AI agents. Predict what should happen, compare against reality, and block the next mutation when they diverge. Pure Python, no service.

A lightweight strategy-driven framework for modular AI tools and automation systems. Core principles: Architecture defines the flow. Strategy controls behavior. Modules provide capabilities. Logs provide feedback. Designed for stability, extensibility, and system-level experimentation.