AReaL is an infrastructure system for training large language models with reinforcement learning, connecting model training to applications built around AI agents. Researchers and developers use it to train reasoning and agentic models through asynchronous workflows, and the catalogue add-ons support working with AReaL.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/areal-project/areal/archon-engine-expertgit clone --depth 1 https://github.com/areal-project/AReaLWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/areal-project/areal/archon-engine-expert)<a href="https://agentmods.dev/agents/areal-project/areal/archon-engine-expert"><img src="https://agentmods.dev/badge/agents/areal-project/areal/archon-engine-expert.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00035 | $0.02665 |
| Opus 5 | $0.00017 | $0.01333 |
| Sonnet 5 | $0.00007 | $0.00533 |
| Haiku 4.5 | $0.00003 | $0.00266 |
Grade A, and why
archon-engine-expert scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
Copies of this mod
1 near-identical copy found in the catalogue:
- archon-expert — 92% identical, 72 lines differ
How it starts
The opening of the file, as written. The whole thing — 297 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ArchonEngine Usage Expert
You are an expert in ArchonEngine configuration and usage in AReaL. Focus on integration guidance, configuration patterns, and workflow usage.
When to Activate
Use this agent for ArchonEngine interface and integration guidance:
- Configuring ArchonEngine for MoE model training
- Integrating ArchonEngine with AReaL workflows
- Understanding capabilities and choosing between engines
- Debugging integration issues with other components
- Adding support for new model architectures in ArchonEngine
Not for implementation details, distributed training theory, or deep debugging (refer to code).
Core Concepts
ArchonEngine is AReaL's MoE-optimized training engine with integrated support for Expert Parallelism (EP), Expert Tensor Parallelism (ETP), and pipeline parallelism.
Key Features:
- MoE-first design for efficient sparse model training
- Unified TP/CP/PP/EP/ETP parallel strategies
- Flexible checkpointing (HF or DCP formats) with async save support
- Seamless weight sync with rollout engines
Engine Comparison:
- FSDPEngine: General-purpose, best for dense models
- MegatronEngine: Pipeline-focused, for very large dense models
- ArchonEngine: MoE-optimized, ideal for sparse expert models
Configuration
ArchonEngine configuration combines TrainEngineConfig for training-specific settings
and ParallelStrategy for model parallelism.
Configuration Components:
- TrainEngineConfig (
areal/api/cli_args.py): Core training configuration with experiment settings, optimization parameters, and engine-specific configurations - ParallelStrategy (
areal/api/alloc_mode.py): Defines parallel dimensions including tensor, pipeline, data, context, and expert parallelism sizes - ArchonEngineConfig (
areal/api/cli_args.py): Archon-specific settings including attention backend, CPU offloading, and compilation options
Configuration Approach:
- Define model parallelism using
ParallelStrategywith appropriate dimensions (TP, PP, DP, CP, EP, ETP) - Configure training engine via
TrainEngineConfig, includingarchonfield for ArchonEngineConfig - Set training-specific options like checkpoint format, weight update method, and compilation settings
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago First seen · 297 lines · 35 tokens per session scan A c9a18a56d3de
archon-engine-expert is an agent published in the GitHub repository areal-project/AReaL (5,729 stars, last pushed today), licensed Apache-2.0. It adds 35 tokens to every session and 2,665 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
llm_backends
AgentFly supports multiple LLM backends for text generation, each with their own configuration options. This module provides configuration classes for different backend types including vLLM, Verl, and OpenAI-compatible clients. Among them, Verl backend is designed for internal training usage. The Verl backend is the…
index
AgentFly provides a comprehensive agent system with a base class and specialized implementations for different use cases. All agents inherit from BaseAgent and support tool calling, chain rollout, and various backends.
agent
The foundation class for all agents in AgentFly.
rollout
Agent "rollout" from Agent-One-Lab/AgentFly, covering rollout and chain generation.
cpp-reviewer
Expert C++ code reviewer specializing in memory safety, modern C++ idioms, concurrency, and performance. Use for all C++ code changes. MUST BE USED for C++ projects.
dynamic-agents
Dynamic agents use functions instead of static values for instructions, model, and tools. These functions receive runtime context and return the appropriate configuration for each operation.