Updates PyTorch AT_DISPATCH_V2 macros to enable uint16/uint32/uint64 type coverage for operators and kernels.
A reference skill for authoring Claude Code subagents: frontmatter rules, triggering examples, and system-prompt structure.
A skill that generates CloudFormation templates for Amazon Bedrock agents, knowledge bases, and guardrails.
Teaches Claude to drive the `vastai` CLI for GPU instances, volumes, serverless endpoints, and billing on Vast.ai.
Performance-minded rules and code patterns for algorithmic art on Canvas 2D — particles, flow fields, noise, fractals and L-systems.
A repo-specific rule set that forces every localization change to land on all five surfaces — UI keys, wiki, mirrored READMEs, language switchers, and locale formatting — in the same PR.
A quiet prompt mentor that scores each prompt by stage and appends one coaching line to every answer in Claude Code or Codex.
A step-by-step playbook for designing, implementing, hardening, and verifying production-grade MCP servers.
A skill for designing reproducible AI agent evaluations with representative datasets, rubrics, graders, regression gates, and a release decision memo.
Routes all codebase exploration to trace-mcp's semantic MCP tools instead of Read/Grep/Glob for higher accuracy and lower token cost.
Forces Claude to use trace-mcp's apply_codemod for any edit pattern repeated 2+ times instead of firing off repetitive Edit calls.
Guides Claude to add or edit MCP server entries in mcptoon's config.json with the correct shape, command rules, and validation loop.
Turns your Git history into a categorized CHANGELOG.md with a Semantic Versioning bump suggestion.
A skill that guides Claude through host-side setup, configuration, builds, migration, and troubleshooting of the Aerovato Container CLI.
The official operating manual that teaches Claude to build widgets and views for a moi web workspace, plus theming, env handling and verification.
Launches the real build on its target runtime and proves six gates — launch, render, input, core loop, a designed outcome, and restart — recording everything in qa/verification.json.
Turns an approved game design into a minimal BUILD_BRIEF plus a runnable candidate, iterated with replayable evidence.
A final quality gate that verifies docs, tests, and security in three checkpoints, then issues a READY or BLOCKED verdict in a written assessment.
Reads a QA code review report, fixes Critical then Warning findings with tests, and produces a fix report for re-verification.
Scaffolds a React + Tailwind + shadcn/ui project and bundles it into a single self-contained HTML artifact for claude.ai.