aka. Agent Skills
Discover skills for AI coding agents. Works with Claude Code, OpenAI Codex, Gemini CLI, Cursor, and more.
Remote-control tmux sessions for interactive CLIs by sending keystrokes and scraping pane output.
Build beautiful, modern websites from codebase or from scratch
Analyzes natural language requests and suggests the most appropriate SiftCoder command to use
Research search trends, keywords, and user intent to inform feature development and content strategy
Vector-based code discovery using LanceDB and Ollama embeddings
Automated code quality validation with configurable checks
Conduct comprehensive market research including competitor analysis, market trends, and user demand signals
Install MCP servers and behavior guides
Reusable capabilities for autonomous multi-file operations
Creates and improves agent skills. Use when creating, updating, or reviewing skills, or when user asks about SKILL.md format, frontmatter fields, skill structure, or skill authoring best practices.
Browser automation for any web task. Use when navigating websites, checking UI, or automating workflows.
Run global performance checks on HTTP endpoints from multiple regions worldwide. Use when users want to check speed, latency, performance, or test endpoints globally.
Multi-agent pub/sub coordination for AI agents working in parallel. Use when: (1) Project has .hydra/ directory initialized, (2) Working with multiple agents or parallel tasks, (3) Need to coordinate, communicate, or share state with other agents, (4) Starting a session and should check what other agents did, (5) Finishing work and should notify other agents. Triggers: multi-agent, coordinate, other agents, parallel agents, emit, broadcast, hydra-mail, check messages, agent coordination.
Build browser automation and web scraping with Playwright on your local machine. Prevents 10 documented errors including CI timeout hangs, extension testing failures, and Ubuntu compatibility issues. Includes stealth mode for anti-bot bypass, authenticated sessions, infinite scroll handling, screenshot/PDF generation, and v1.57 Speedboard performance analysis. Use when: automating browsers, scraping protected sites, testing with real IPs, bypassing bot detection, generating screenshots/PDFs, or troubleshooting "target closed", "page.pause() hangs CI", "permission prompts block tests", or "Ubuntu 25.10 installation" errors.
MCP tools quick reference for Giant Swarm operations
Giant Swarm GitHub label taxonomy and usage guidelines
Quality standards and best practices for Giant Swarm runbooks. Use when creating or migrating runbooks.
To keep in mind when working with Kubernetes
Best practices when working with Flux CD at Giant Swarm
The Backstage developer portal