Skill Claude CodeCodex
Use this skill when designing, generating an MVP blueprint for, auditing, refactoring, or explaining an agentic harness for any domain. Covers provider-neutral agent architecture for OpenAI, Anthropic, and OpenAI-compatible APIs: agent loops, tool design, permissions, system prompts, planning, goals, context…
Skill Claude CodeCodex
Ultra-compressed communication mode. Cuts output tokens 65% (measured) by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, wenyan-ultra. Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens"…
Skill Codex
Use when performing any file-changing development task in a repository, including code, tests, documentation, configuration, schemas, generated artifacts, implementation, fixes, refactors, finishing, or verification.
Skill Claude CodeCodex
Behavioral guidelines to reduce common LLM coding mistakes. Use when writing, reviewing, or refactoring code to avoid overcomplication, make surgical changes, surface assumptions, and define verifiable success criteria.
Skill Claude CodeCodex
Karpathy's LLM Wiki: build/query interlinked markdown KB.
Skill Claude Code
Use when starting work in a project repository that has no agent harness or an incomplete one (no root AGENTS.md canon, no .harness/ directory, no gates, no harness.lock), when a SessionStart hook reports a missing or outdated harness, or when the owner asks to set a project up for agent-driven development. Detect…
Skill Claude CodeCodex needs its repo
Use when finishing project-scoped work that changed project state or produced a completed audit, diagnosis, investigation, or decision, including tasks with or without a tracker.
Skill Claude CodeCodex
A library of patterns for turning vague coding requests into clear prompts for Claude Code. It covers common development phases and work such as onboarding, incidents, migrations, releases, and reviews.
Skill Claude CodeCodex
Use when writing an implementation plan for a multi-step coding task before execution: bite-sized tasks, exact paths, complete code, verification steps per task.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: