Openmythos

Use when working with the OpenMythos codebase (Kye Gomez's PyTorch Recurrent-Depth Transformer): open_mythos/main.py, MythosConfig, OpenMythos, RecurrentBlock, LTIInjection, ACTHalting, MoEFFN, MLAttention, GQAttention, LoRAAdapter, mythos_1b..mythos_1t variants, Prelude/Recurrent/Coda layout, h=A*h+B*e+Transformer(h,e) update, loop-index embedding, ACT remainder trick, FineWeb-Edu looped training, or looped-training symptoms (loss spikes at a reproducible step, residual explosion, overthinking). Also when asked for 'Mythos mode'. A filename or symbol match is enough. Not for generic transform

9 stars 4 forksAI, LLMs & Agents

Install

skills CLI (Claude Code, Cursor, Codex and more)

npx skills add SarthakDz/OpenMythos-Skill --skill openmythos

Manual install

git clone --depth 1 https://github.com/SarthakDz/OpenMythos-Skill.git /tmp/OpenMythos-Skill
mkdir -p ~/.claude/skills
cp -r /tmp/OpenMythos-Skill/. ~/.claude/skills/openmythos

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# OpenMythos

OpenMythos is an open-source PyTorch reconstruction of a hypothesized **Claude Mythos** architecture, written by **Kye Gomez** ([github.com/kyegomez/OpenMythos](https://github.com/kyegomez/OpenMythos), MIT license). It implements a **Recurrent-Depth Transformer (RDT)** with three stages — **Prelude** (standard transformer blocks, run once), a **Recurrent Block** (one TransformerBlock looped up to `max_loop_iters` times with input injection at every step), and a **Coda** (standard transformer blocks, run once). Attention is switchable between GQA and MLA; the FFN inside the recurrent block is a fine-grained MoE with always-on shared experts.

The project is an *independent, theoretical* reconstruction. It is not affiliated with Anthropic. The README is careful with language like "suspected", "likely", and "most probable class of solution", and so is this skill — don't claim this is what Anthropic actually does internally. If the user conflates OpenMythos with real Claude internals, gently correct them.

This skill turns Claude into a careful senior engineer who knows this specific repo. That's the entire job. (There is an optional experimental appendix at the bottom for users who want Claude to roleplay reasoning in the RDT's Prelude → Loop → Coda shape, but it is off by default.)

## The repo at a glance

Read the full skill on GitHub

Details

Skill name: openmythos
Repository: SarthakDz/OpenMythos-Skill
License: MIT
Repo updated: 4 October 2026

FAQ

What does the Openmythos skill do?

How do I install the Openmythos skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add SarthakDz/OpenMythos-Skill --skill openmythos. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Openmythos popular?

It lives in SarthakDz/OpenMythos-Skill, which has 9 GitHub stars and 4 forks.

Related skills

200 Agent

HeiGeAi/200-agent

Route a task to one of 200 expert AI playbooks across software, AI, design, writing, data, product, marketing, business, research, professional advice, productivity, operations, and industry domains. Use when the user wants specialist-quality guidance, asks which expert role fits a task, names a 200-agent slug, or needs a cross-domain expert workflow.

AI, LLMs & Agents

Openclaw Genpark Compliance Shield

alphaparkinc/openclaw-genpark-compliance-shield

OpenClaw GenPark Compliance Shield - Agent autonomous policy and differential privacy limiter with RBAC/ABAC rules and Laplace noise mechanism.

AI, LLMs & Agents

AI Film Studio

realaman90/ai-film-skills

AI film production studio with pre-built scripts: analyze inspiration/reference videos (Gemini), write script + storyboard, generate stills (GPT Image 2.5 Sunburst by default; Flare for previs; GPT Image 2 / Seedream 5 / Nano Banana as fallbacks), video (Gemini Omni 1.1 Flash, FLUX 3 Video, Seedance 2.5/2.0, LTX 2.5/2.3; Veo 3.1 fallback), audio (ElevenLabs), and assemble the final cut with HyperFrames (HTML→MP4; replaced Remotion) or ffmpeg. Use when: creating short films, ads, product promos, brand films, social reels, storyboards, video scenes, voiceovers, music, sound effects, or any AI-po

AI, LLMs & Agents

Guide Mentor

alphaparkinc/genpark-skill-guide-mentor

Guide Mentor — Proactive project management agent that provides high-stakes guidance, iterative feedback, and strategic oversight for complex, long-running projects.

AI, LLMs & Agents

Ambition

MontySinger/ambition

Turn a rough request into a more useful assignment by reconsidering its outcome, choosing concrete methods and evidence, and defining how to check the result. Use when asked to improve a prompt or brief, make an approach more ambitious, or rethink work that is not helping. Can prepare a prompt or improve and execute an authorized task with other available skills.

AI, LLMs & Agents

Prompt

gpaiva00/pt-prompt

Use when translating Brazilian Portuguese messages into English prompts for Claude Code — preserves intent, clarifies ambiguity, and surfaces urgency signals that colloquial Portuguese would lose.

AI, LLMs & Agents

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.