The Agent Skills Directory

[SAFE]: The skill provides a framework for evaluating documentation quality. It is restricted to using Read, Glob, and Grep for data ingestion and analysis, ensuring that it cannot modify the file system or communicate over the network.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection (Category 8) as it ingests and processes external SKILL.md files.
Ingestion points: Local SKILL.md files are read using the Read tool based on user-provided paths.
Boundary markers: Absent. The skill does not explicitly define markers to separate analyzed content from the agent's instructions.
Capability inventory: The skill's allowed tools are limited to information gathering (Read, Glob, Grep). It lacks capabilities for code execution, system modification, or network access.
Sanitization: Content is analyzed as raw text without specific sanitization steps.

rate-skill