OpenMythosAgent — Development skill for Claude Code
OpenCode agent + cloned source from kyegomez/OpenMythos — Recurrent-Depth Transformer (RDT) with Prelude/Recurrent/Coda, MLA, MoE, LTI-stable injection, ACT halting.
How to install OpenMythosAgent
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open xanstomper/OpenMythosAgent and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What OpenMythosAgent does
OpenCode agent + cloned source from kyegomez/OpenMythos — Recurrent-Depth Transformer (RDT) with Prelude/Recurrent/Coda, MLA, MoE, LTI-stable injection, ACT halting. PyTorch.
Alternatives in Development
- OpenMythos — A theoretical reconstruction of the Claude Mythos architecture, built from first principles using the availabl 14.8k ★
- Promote Public Release — Build, refine, and publish cumulative public release notes, then promote the current alpha prerelease to stabl 1.6k ★
- Dao Code — Open-source TypeScript terminal coding agent for DeepSeek-V4 — builds on DeepSeek's strong price-performance a 1.3k ★
README
╔══════════════════════════════════════════════╗ ║ ║ ║ ███╗ ███╗██╗ ██╗████████╗██╗ ██╗ ║ ║ ████╗ ████║╚██╗ ██╔╝╚══██╔══╝██║ ██║ ║ ║ ██╔████╔██║ ╚████╔╝ ██║ ███████║ ║ ║ ██║╚██╔╝██║ ╚██╔╝ ██║ ██╔══██║ ║ ║ ██║ ╚═╝ ██║ ██║ ██║ ██║ ██║ ║ ║ ╚═╝ ╚═╝ ╚═╝ ╚═╝ ╚═╝ ╚═╝ ║ ║ ║ ║ Recurrent-Depth Transformer — Agent ║ ╚══════════════════════════════════════════════╝
**OpenCode agent + source** from [`kyegomez/OpenMythos`](https://github.com/kyegomez/OpenMythos) — the open-source PyTorch implementation of the **Recurrent-Depth Transformer (RDT)**, the theoretical architecture behind Claude Mythos.
This repository packages the full OpenMythos source into an **OpenCode agent** that applies RDT-style reasoning — prelude encoding, recurrent latent iteration, LTI-stable state updates, MoE routing, ACT halting, and coda refinement — to any base model.
Table of Contents
Related Skills
OpenMythos Skill
OpenMythos Claude Skill: a lightweight add-on that makes Claude act like a senior engineer who deeply understa
Prelude
Prelude is the open standard for machine-readable project context — enabling LLMs, agents, and teams to unders
NarrateClaude
Hands-free on-device voice loop for macOS: Apple SFSpeechRecognizer + cloned-voice TTS, continuous listening,
Clean Instantiate
Clean a freshly-cloned client repo of canon operator/trust identity BEFORE /ecosystem-init (clear-then-bootstr
Agent Repo Preflight
Read-only preflight for cloned repos: inventories AGENTS.md, CLAUDE.md, .cursorrules and MCP/hook configs, and
Deep Rl
Deep reinforcement learning - DQN/Rainbow/R2D2/Agent57/BBF, PPO/TRPO/GRPO, SAC/TD3/REDQ/DroQ/CrossQ, DreamerV3
Related Agents
Knowledge Auditor
Independently audits a distilled library knowledge bundle against the library's own cloned source at the pinne
Security Architecture
Boucle implements a defense-in-depth security model to protect against prompt injection attacks and maintain t
Database Attacker
Delegates to this agent when the user wants database-specific offensive testing on an authorized target — SQL