The Agent Skills Directory

PROMPT_INJECTION (HIGH): The skill's configuration creates a high-risk surface for Indirect Prompt Injection because it combines data ingestion with file-modification capabilities.
Ingestion points: The frontmatter enables Read, Glob, and Grep tools, which are used to ingest content from the user's workspace files such as the referenced database selection or schema design documents.
Boundary markers: The instructions do not define any delimiters or provide a security context to help the agent distinguish between its own system-level instructions and untrusted data contained in the files it reads.
Capability inventory: The frontmatter enables Write and Edit tools, allowing the agent to perform side effects on the filesystem, which could be leveraged by an attacker to overwrite sensitive code.
Sanitization: The skill does not perform any validation, sanitization, or filtering of the content retrieved from external files before the agent processes it.

database-design