A collection of Jupyter notebooks and Python examples for building with the Claude API.
Skill Creation and Aider Skills
2 skills match Skill Creation plus Aider. Clear filters
The Agent Skill Source directory is architected into three primary discovery layers: Outcome-based Categories (what you want to achieve), Runtime Compatibility (where the agent runs), and Execution Strategy (how the skill is processed). To find the most reliable capabilities, use the Highest Rated sort and filter by Verified trust levels.
Search results
Master prompt evaluation techniques for AI models using Anthropic API, Workbench, and promptfoo across nine comprehensive lessons.
Other ways into the directory
Filters narrow the list. These pages start you somewhere specific instead, with the context that explains what you are looking at.
Browse by runtime
Popular tags
Cannot find what you need?
Submit the GitHub repository and we will parse the SKILL.md, run the trust checks, and publish it. You keep the credit as the author.
Submit a skill