OpenMythos

Codebase-expert prompt for Kye Gomez's OpenMythos looped-transformer PyTorch repo.

Tested · Works

Test report

Verdict
Tested · Works
Score
9.6/10
Tested
Jul 16, 2026
Environment
Claude Code 2.x (agent harness)
Upstream re-checked
Aug 10, 2026 · fbe2689

Fed it a KV-cache-error-on-decode-step-2 question: it went straight to the actual repo mechanism (per-loop-depth cache keys + the kv_cache-vs-break invariant) instead of the generic cache-debugging checklist a baseline answer produced.

Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works

  • Installs cleanly 5/5
  • Triggers reliably 5/5
  • Output vs. baseline 9/10
  • Docs & honesty 5/5

What OpenMythos does

Turns Claude into a repo-specific senior engineer for OpenMythos, an independent PyTorch reconstruction of a hypothesized looped/recurrent-depth transformer architecture. Triggers on filenames, imports, symbol names (MythosConfig, RecurrentBlock, LTIInjection), variant helpers, and looped-training-specific debugging symptoms (spectral-radius drift, KV-cache errors on decode step 2+, overthinking degradation).

How to install OpenMythos

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

Skills live in ~/.claude/skills/ (global) or .claude/skills/ (per-project). Restart Claude Code after installing.

Commands — how to trigger OpenMythos

  • /openmythos Codebase-expert prompt for Kye Gomez's OpenMythos looped-transformer PyTorch repo.

It also activates on plain-language prompts like these:

  • Decode step 2 throws a KeyError in the recurrent block's KV cache
  • Why is my mythos_1b inference showing spectral-radius drift?
  • Debug this overthinking degradation bug in my looped transformer run

Frequently asked questions

Is the OpenMythos skill free?
Yes. The skill itself is free from SarthakDz/OpenMythos-Skill. SkillProof publishes the install command and an independent test verdict at no cost.
Does OpenMythos work with Claude Code?
We tested it with Claude Code 2.x (agent harness) on Jul 16, 2026. Verdict: Tested · Works. Fed it a KV-cache-error-on-decode-step-2 question: it went straight to the actual repo mechanism (per-loop-depth cache keys + the kv_cache-vs-break invariant) instead of the generic cache-debugging checklist a baseline answer produced.
What is the OpenMythos SkillProof Score?
9.6/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 9/10, docs & honesty 5/5.
How do I install OpenMythos?
Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
Can I use OpenMythos with Cursor, Copilot, Gemini CLI, Codex or other AI tools?
The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.