Skip to content

Repository files navigation

Defensive Agent Skills

This repository contains benign, non-executable skill definitions created as defensive placeholders for agent-skill hallucination research.

Each skill is a text-only SKILL.md with safety boundaries and review workflows. The repository intentionally does not include scripts, hooks, MCP configurations, binaries, credential handling, network callers, or code that performs actions on a user's system.

These skills are intended to reduce ambiguity around commonly hallucinated skill names while providing harmless guidance if an agent selects one of them.

About

skills for agent

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors