The Agent Skills Directory

Indirect Prompt Injection (LOW): The skill is designed to analyze untrusted content from project files, which could contain malicious instructions meant to subvert the agent's logic. \n
Ingestion points: The skill ingests data from README.md, project-brief.md, CLAUDE.md, and various files in specs/ or issues/ via the Read and Grep tools. \n
Boundary markers: No delimiters or instructions are specified to help the agent distinguish between its instructions and the data being validated. \n
Capability inventory: The potential impact is limited because the skill only allows Read, Glob, and Grep. It cannot perform write operations, execute code, or access the network. \n
Sanitization: The skill does not perform any sanitization or validation on the content of the documentation it reads.

validate-idea