My personal skills repository for LLM agents
Response formatting rules for swag2mcp MCP tools. Ensures consistent human-readable markdown, enforces pagination, colorizes HTTP methods, and structures schemas, headers, and errors for all swag2mcp tool responses.
📥 Download
All models behave differently. While some models may automatically discover and load skills, others require explicit instructions. To ensure consistent and reliable behavior across different LLMs, we strongly recommend explicitly loading skills at the start of each session.
Begin every session by loading the skills you need. This guarantees that the model has access to all instructions, templates, and resources regardless of its default behavior.
Use these prompts to control skill loading:
| Action | Prompt |
|---|---|
| List all available skills | Show all available skills |
| Load all skills | Load all skills |
| Load a specific skill | Load the [skill-name] skill (e.g., Load the swag2mcp skill) |
- Consistency – Ensures the same behavior across different models and sessions
- Control – You decide exactly which skills are active
- Transparency – Verify what's available before starting your task
- Reliability – Avoids assumptions about model auto-discovery capabilities