Scripting skills

This skill executes terminal commands safely and incrementally, validating each step to reduce risk and ensure reliable outcomes.

This skill enables building and coordinating recursive multi-agent systems with conflict-free parallel task execution and automatic task decomposition.

This skill helps migrate CUDA code to the MUSA platform by transforming CUDA kernels and APIs with text-based mapping.

This skill identifies flaky GitHub Actions steps by aggregating run outcomes and flags instability across repos, workflows, jobs, and steps.

This skill audits GitHub Actions workflows for cache hardening, identifying risky patterns that enable poisoning, leakage, or stale cache issues.

This skill provides useful helper functions for the agent to streamline tasks, improve reliability, and accelerate automation workflows.

This skill audits GitHub Actions workflows to identify hardening gaps, flag missing timeouts and permissions, and prevent risky defaults from reaching CI.

This skill searches Google via SerpAPI for web, news, and local results with country and language targeting.

This skill audits GitHub Actions workflows for OIDC hardening gaps, flagging missing id-token permissions, static secrets, and floating refs to improve cloud

This skill extracts PDF content to Markdown using MinerU, including formulas, tables, and OCR, for easy offline and online parsing.

This skill creates a Google Classroom course and invites students using gws-classroom, enabling streamlined setup and enrollment.

This skill enhances Gmail management with AI triage, priority scoring, smart replies, and thread summaries to boost inbox zero.

This skill generates polished, branded client intake PDFs for pet sitters, including multi-pet forms, color themes, and service-specific templates.

This skill screens FTSE 350 stocks, analyzes indicators, and outputs risk-aware trade recommendations with sentiment and backtesting insights.

This skill helps you send rich Feishu 2.0 schema cards with double JSON stringification, ensuring reliable rendering across locales.

This skill evaluates AI agent performance across reasoning, retrieval, creation, execution, and orchestration, saving results for history review and progress

This skill helps your OpenClaw agent manage long-term memory with automatic archiving and skill extraction for reusable capabilities.

This skill operates SNU Canvas via the CLI bridge to inspect config, list courses and files, and troubleshoot API issues safely.

This skill helps you find Taiwan surf spots with real-time tides, wind, typhoon updates, sunrise/sunset, and nearby parking via Telegram, LINE, or iMessage.

This skill detects prompt injection attempts in emails, blocks dangerous instructions, and prompts user confirmation before executing any action.

This skill enables voice message transcription and synthesis using faster-whisper and Edge TTS to deliver replies.