JSPM

Agent Model Layer — One-line LLM access with built-in memory. 29 providers, 18 embeddings, zero lock-in.

Fresh-context iterative loops for Claude Code — autonomous task completion with independent verification

YUAN Agent Core Runtime

Helixent is a blue rabbit that writes code. It includes an Agent Loop, a Coding Agent, and a nice CLI.

One-command OpenCode installer for Agent Loop

Unified AI creation engine — text, image, video, audio across all providers

Agent run-loop primitive over the Tangle sandbox bridge — runs an AgentProfile against a brief with criterion gates, budget caps, and streaming events. Includes TangleToolProvider for Pi tool integration.

Middleware framework for agent control loops with semantic drift detection and geometric control

Embeddable TypeScript evolutionary loop: Zod Seeds, Wonder/Reflect, Executor/Evaluator, ontology convergence, optional MCP tool shapes — use with any agent backend or job worker (Node ≥22).

Universal loop controller for AI agents and async workflows. Control iterative execution, prevent infinite loops, and manage agent cycles with precision.

Loop-keeper for AI coding agents — plans into phases, verifies each, auto-continues until done.

Structured workflow kata for Claude Code — modes, phases, tasks, and stop hooks for disciplined AI coding sessions

Level 1 Foundation — Personal AI Architecture. Generic, user-owned AI runtime with zero lock-in.

Constraint-learning runtime for iterative agent loops. Agents that structurally learn what NOT to try again.

Structured workflow kata for Claude Code — modes, phases, tasks, and stop hooks for disciplined AI coding sessions

AgentRuler SDK — AgentLoop core with Node.js backend adapter, permission system, and full toolbar settings API.