A set of rules for preventing an AI agent from repeating failed tool calls or getting stuck in a loop. It requires the agent to inspect errors, use new information, change approach, and stop after repeated failures.
A working mode that balances careful thinking with practical execution. It calls for analyzing the request, making a plan, carrying it out, checking the result, and explaining what was done.
A full code-audit workflow that looks for logic, interface, translation, security, and performance problems. It requires every relevant file to be read completely and reports findings without changing the code.
A set of coding and repository conventions covering Python, Telegram bots, project structure, typing, logging, documentation, Docker, and deployment files.
A coding skill that guides an agent toward careful, minimal changes and clear communication. It emphasizes checking assumptions, defining success, and avoiding unnecessary complexity.
A meta-skill for creating new agent skills: written instructions that guide an agent through a repeatable task.
★not rated 2 1mo agoA42 tokens
originalMIT
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: