Discover and explore top open-source AI tools and projects—updated daily.
rmyndharisAgent skills for extending AI agent capabilities
Top 80.2% on SourcePulse
This repository offers a curated collection of over 300 specialized agent skills for Google Antigravity, ported from the Claude Code Agents ecosystem. It aims to equip Antigravity agents with repeatable workflows, domain expertise, and specialized tools across software development, operations, security, and business domains, thereby enhancing their capability to perform complex tasks autonomously. The target audience includes users of Google Antigravity seeking to extend their agent's functionality with pre-built, domain-specific capabilities.
How It Works
Skills are structured as directory-based packages, each containing instructions and resources for a specific task. These are derived from Claude Code's Domain Skills, Specialist Agents, and Commands & Workflows, unified into the Antigravity Skill format. Upon session start, Antigravity loads the metadata (name and description) of all installed skills. The agent then automatically activates a skill when a user's request semantically matches its description, enabling context-driven task execution and automation.
Quick Start & Requirements
npx @rmyndharis/antigravity-skills <command> (e.g., search, list, install).npx. Requires a Google Antigravity agent setup.npx CLI commands.Highlighted Details
CATALOG.md (human-readable), catalog.json (machine-readable), bundles.json (curated sets), and aliases.json (short names).kubernetes), or bundles (e.g., core-dev, security-core).Maintenance & Community
The repository is a port of original Claude Code Agents content. No specific details on active maintainers, community channels (like Discord/Slack), or a public roadmap are provided in the README.
Licensing & Compatibility
The repository is licensed under the MIT License. The README notes that the content is "Original content © Claude Code Agents. Ported to Antigravity Skills." This MIT license applies to the packaging and distribution of these skills within this repository.
Limitations & Caveats
Installing all skills (install --all) is explicitly discouraged due to increased token usage and potential for irrelevant skill auto-activation. Manual installation via cp requires restarting the agent session for skills to be detected. Authoring guidelines suggest using narrow descriptions and "Do not use" clauses to fine-tune skill activation. Skills impacting terminals or infrastructure require adherence to safety expectations outlined in SECURITY.md.
1 month ago
Inactive