The Agent Skills Directory

[PROMPT_INJECTION]: The skill's refactoring workflow relies on fetching and following instructions from an external llms.txt file, creating a surface for indirect prompt injection.
Ingestion points: rules/react-util-usage.md directs the agent to access https://raw.githubusercontent.com/mj-studio-library/react-util/master/llms.txt.
Boundary markers: Absent. The skill does not provide delimiters or instructions for the agent to ignore potentially malicious content within the fetched documentation.
Capability inventory: The agent has the capability to generate, refactor, and implement React utility logic based on the external content.
Sanitization: Absent. External documentation is treated as a direct implementation guide without validation or filtering.
[EXTERNAL_DOWNLOADS]: The skill references and manages external resources from well-known services.
Evidence: Instructs the agent to install @mj-studio/react-util via package managers and fetch documentation from GitHub repositories.
Source: Resources are hosted on GitHub and NPM, which are well-known technology platforms.

react-skills