CLI Tools
Command line utilities
cate-cli
by 0-AI-UG
Drive Cate browser, terminal, editor, panel, and coding-agent orchestration surfaces from a Cate terminal. Browser page automation uses native agent-browser command syntax.
square-post
by binance
Use when the user wants to publish new content to Binance Square — short text, multi-image posts (up to 4), long-form articles with an optional cover, or videos with an auto-generated cover frame. Trigger on direct phrasings like "post to Square", "publish to Binance Square", "发广场", "发布到广场", and on near-miss intents where the user clearly wants to share or publish content on Square even without naming the skill: "share this analysis on Square", "把这篇文章发出去", "发个动态", "把这个视频上传到广场", "publish my chart to Square as an article". Also use when the user provides media (images, video) plus a caption and asks to push it to Square, or asks to turn a draft into a Square article. Do not use for reading, searching, commenting, liking, editing, deleting, scheduling, or managing existing Square posts — this skill only creates new posts.
security-scan
by affaan-m
Scan your Claude Code configuration (.claude/ directory) for security vulnerabilities, misconfigurations, and injection risks using AgentShield. Checks CLAUDE.md, settings.json, MCP servers, hooks, and agent definitions.
verification-loop
by affaan-m
"A comprehensive verification system for Claude Code sessions."
skypilot-multi-cloud-orchestration
by Orchestra-Research
Multi-cloud orchestration for ML workloads with automatic cost optimization. Use when you need to run training or batch jobs across multiple clouds, leverage spot instances with auto-recovery, or optimize GPU costs across providers.
training-llms-megatron
by Orchestra-Research
Trains large language models (2B-462B parameters) using NVIDIA Megatron-Core with advanced parallelism strategies. Use when training models >1B parameters, need maximum GPU efficiency (47% MFU on H100), or require tensor/pipeline/sequence/context/expert parallelism. Production-ready framework used for Nemotron, LLaMA, DeepSeek.
optimizing-attention-flash
by Orchestra-Research
Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction. Use when training/running transformers with long sequences (>512 tokens), encountering GPU memory issues with attention, or need faster inference. Supports PyTorch native SDPA, flash-attn library, H100 FP8, and sliding window attention.
lambda-labs-gpu-cloud
by Orchestra-Research
Reserved and on-demand GPU cloud instances for ML training and inference. Use when you need dedicated GPU instances with simple SSH access, persistent filesystems, or high-performance multi-node clusters for large-scale training.
evaluating-code-models
by Orchestra-Research
Evaluates code generation models across HumanEval, MBPP, MultiPL-E, and 15+ benchmarks with pass@k metrics. Use when benchmarking code models, comparing coding abilities, testing multi-language support, or measuring code generation quality. Industry standard from BigCode Project used by HuggingFace leaderboards.
evaluating-llms-harness
by Orchestra-Research
Evaluates LLMs across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use when benchmarking model quality, comparing models, reporting academic results, or tracking training progress. Industry standard used by EleutherAI, HuggingFace, and major labs. Supports HuggingFace, vLLM, APIs.
gguf-quantization
by Orchestra-Research
GGUF format and llama.cpp quantization for efficient CPU/GPU inference. Use when deploying models on consumer hardware, Apple Silicon, or when needing flexible quantization from 2-8 bit without GPU requirements.
distributed-llm-pretraining-torchtitan
by Orchestra-Research
Provides PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use when pretraining Llama 3.1, DeepSeek V3, or custom models at scale from 8 to 512+ GPUs with Float8, torch.compile, and distributed checkpointing.
cmux-capture
by manaflow-ai
Screenshot or record a cmux window from the CLI, for a pull request, a bug report, or to see what the app actually looks like right now. Use when a change is visual, when a drag or an animation cannot be described in text, or when a reviewer needs evidence rather than a claim.
minimum-viable
by ferrumclaudepilgrim
Before building anything, justify the tool choice and complexity level. Could a shell script do this? Does it need Node.js? A full application? Prevents over-engineering by requiring explicit justification before choosing a heavier approach.
ra-video-download
by Pluviobyte
"Download source video or audio from Douyin, YouTube, Bilibili, Twitter/X, Xiaohongshu, and other yt-dlp-supported URLs into the content-creation workspace. Use when the user says 下载视频, 下载音频, 保存这个链接, 把源视频拉下来, or when another video workflow needs a local source file before transcription, frame analysis, editing, reference replication, or archiving."
openalex
by brycewang-stanford
Use when the user asks about academic literature, research papers, scholarly works, authors, citations, institutions, journals, or any academic metadata. Trigger when users want to search for papers, find author profiles, track citations, discover related works, or explore academic topics. Also use when users mention DOIs, ORCIDs, h-index, publication venues, or research metrics.
address-sanitizer
by trailofbits
AddressSanitizer detects memory errors during fuzzing. Use when fuzzing C/C++ code to find buffer overflows and use-after-free bugs.
agent-recall
by Goldentrii
Persistent compounding memory for AI agents. 5 default MCP tools: session_start, session_end, remember, recall, check. Full surface (18 tools) available with --full flag. Two-verb model: inhale (session_start) and exhale (session_end). Correction-first memory with decision trail tracking, watch_for warnings, palace rooms with salience scoring, cross-project insight matching, same-day journal merging, ambient recall hooks. Local markdown only. Zero cloud, zero telemetry, Obsidian-compatible. Optional Supabase backend: when configured via ar setup supabase, recall() uses pgvector cosine similarity on OpenAI/Voyage embeddings instead of keyword search — same API, semantic understanding. Gracefully degrades to local search if not configured.
linear-projects
by Finesssee
Manage Linear projects. Use when listing, creating, updating, or viewing projects.
linear-roadmaps
by Finesssee
View Linear roadmaps. Use when viewing roadmap planning.
git-workflow
by bobmatnyc
Essential Git patterns for effective version control, eliminating redundant Git guidance per agent.
hyva-compile-tailwind-css
by hyva-themes
Compile Tailwind CSS for Hyvä themes in Magento 2. This skill should be used when the user wants to build styles, generate CSS, compile Tailwind, run Tailwind, or create production/development stylesheets for a Hyvä theme. Triggers on phrases like "compile tailwind", "build styles", "generate css", "run tailwind", "build css", or "npm build for theme".
bats-testing-patterns
by rmyndharis
Master Bash Automated Testing System (Bats) for comprehensive shell script testing. Use when writing tests for shell scripts, CI/CD pipelines, or requiring test-driven development of shell utilities.
trading-signal
by binance
Per-trade smart-money signals — each result is a discrete buy or sell event from a tracked smart-money wallet, with trigger price, current price, max gain since trigger, and exit rate. BSC and Solana only. Use for: "smart money buy signal on $X", "any whale just bought $Y", "alpha signals in the last hour", "copy-trade-worthy signals", "trigger price and max gain on these trades", "on-chain trading signals from smart money".