Reading
AgentLoom is a self-hosted multi-agent workflow runner compatible with Claude Code's dynamic workflow API. Built on OpenCode, it supports deterministic AI workflow execution, sub-agent fan-out, custom tool registration, prompt caching optimization, and features a real-time Web UI that displays tokens and tool calls for each running agent.
Section 01
AgentLoom is a self-hosted multi-agent workflow runner compatible with Claude Code's dynamic workflow API. Built on OpenCode, it supports deterministic AI workflow execution, sub-agent fan-out, custom tool registration, prompt caching optimization, and features a real-time Web UI that displays tokens and tool calls for each running agent.
Section 02
Section 03
With the rapid advancement of Large Language Model (LLM) capabilities, AI system architectures based on Agents are emerging. From single-agent conversations to multi-agent collaboration, from simple task execution to complex workflow orchestration, developers need more powerful tools to manage and coordinate these AI workflows.
AgentLoom was born to address this need. It is a self-hosted multi-agent workflow runner whose API design is fully compatible with Claude Code's dynamic workflow features. This means developers can run workflow scripts that originally relied on the Claude Code client in a local environment, gaining greater flexibility and control.
The core positioning of the project is: to provide a deterministic, reproducible, and observable AI workflow execution engine, enabling developers to build complex agent collaboration systems while maintaining full control over the execution process.
AgentLoom has a direct lineage with Claude Code's dynamic workflow feature. As an AI programming assistant launched by Anthropic, Claude Code has a built-in powerful workflow system that supports AI to automatically generate and execute multi-step tasks through natural language descriptions.
AgentLoom implements the same API surface, including:
agent(), pipeline(), parallel(), phase(), log(), argsmeta = { name, description, phases }pipeline() stages use streaming processing instead of waiting for all to completeThe main differences between the two are:
| Feature | Claude Code Dynamic Workflow | AgentLoom |
Section 04
Original Author and Source
Original Author/Maintainer: vblanco20-1
Source Platform: GitHub
Original Title: AgentLoom
Original Link: https://github.com/vblanco20-1/AgentLoom
Source Publication/Update Time: 2026-05-28T18:45:11Z
Original Author/Maintainer: vblanco20-1
Source Platform: GitHub
Original Title: AgentLoom: Prototype agentic orchestrator for deterministic AI workflows
Original Link: https://github.com/vblanco20-1/AgentLoom
Publication Time: May 28, 2026
Project Background and Positioning
With the rapid advancement of Large Language Model (LLM) capabilities, AI system architectures based on Agents are emerging. From single-agent conversations to multi-agent collaboration, from simple task execution to complex workflow orchestration, developers need more powerful tools to manage and coordinate these AI workflows.
AgentLoom was born to address this need. It is a self-hosted multi-agent workflow runner whose API design is fully compatible with Claude Code's dynamic workflow features. This means developers can run workflow scripts that originally relied on the Claude Code client in a local environment, gaining greater flexibility and control.
The core positioning of the project is: to provide a deterministic, reproducible, and observable AI workflow execution engine, enabling developers to build complex agent collaboration systems while maintaining full control over the execution process.
Relationship with Claude Code's Dynamic Workflow
AgentLoom has a direct lineage with Claude Code's dynamic workflow feature. As an AI programming assistant launched by Anthropic, Claude Code has a built-in powerful workflow system that supports AI to automatically generate and execute multi-step tasks through natural language descriptions.
AgentLoom implements the same API surface, including:
agent(), pipeline(), parallel(), phase(), log(), argsmeta = { name, description, phases }pipeline() stages use streaming processing instead of waiting for all to completeThe main differences between the two are:
| Feature | Claude Code Dynamic Workflow | AgentLoom |
Keep going with more reads from the same topic.
Nornir MCP Server is an enterprise-level server based on the Model Context Protocol (MCP). It seamlessly integrates large language models (such as Claude) with the Nornir network automation framework, supporting natural language orchestration for multi-vendor network devices (Cisco, Arista, Juniper, etc.), and providing production-grade features like a dual-engine architecture (NAPALM + Netmiko), intelligent filtering, and a secure sandbox.
Bibliothèque Française LLM is a structured indexing and annotation project for French public domain literature designed specifically for large language models (LLMs). It integrates multiple authoritative sources such as DraCor, Common Corpus, and Wikisource, providing metadata indexing categorized by genre, author, and era, as well as in-depth annotations for dramatic texts (including characters, lines, stage directions, etc.). Its aim is to enable LLMs to efficiently read and understand classic French literary works.
Splinter is a minimalist, high-performance key-value (KV) and vector storage system enabling zero-latency inter-process communication via shared memory and atomic operations. With only 766 lines of core code, it supports millions of operations per second and 768-dimensional vector storage, offering a new architectural approach for local LLM inference and data-intensive applications.
libmlxforge is an embeddable MLX large language model (LLM) inference engine designed specifically for Apple Silicon. It provides a unified C ABI interface, supports calls from Node.js, Swift, and Rust, and features continuous batching, streaming output, JSON-constrained structured output, and embedding vector generation.